A qualitative exploration of factors influencing medical staffs’ decision-making around nutrition prescription after colorectal surgery

Enhanced Recovery After Surgery (ERAS) guidelines assist clinicians in making informed decisions regarding optimal perioperative care among colorectal patients [1, 2]. Early oral feeding (EOF), defined as initiating liquids and solids within 24 h after colorectal surgery [1‐3], is an essential component of ERAS that has proven to be safe and beneficial [4, 5]. However, recent studies have demonstrated variable and often poor adherence to EOF recommendations and frequent use of nutritionally inadequate diets (i.e. clear liquids) among patients that undergo gastrointestinal surgery [6‐8]. In addition to missing out on the benefits associated with EOF, delayed feeding and extended use of liquid only diets can increase the risk of developing or exacerbating protein-energy malnutrition (PEM) [9]. Among surgical patients, PEM has been associated with increased incidence of complications [10], longer lengths of stay [11, 12] and higher rates of hospital readmission [11]. As such, efforts to improve postoperative nutrition prescription practices are needed to optimize patient and healthcare outcomes.

Translating evidence-based guidelines into practice often requires a multifaceted approach, including education, training, and behaviour change techniques [13]. However, an important aspect of translational work is to understand the perceptions of those who will be involved and the environment where the change will occur before attempting to enact change [14]. Medical staff are influential decision makers of any ERAS program and are responsible for prescribing patients’ diets after surgery. As such, previous studies have explored surgeons’ knowledge, attitudes, and behaviours towards perioperative nutritional practices [15‐17]. While this work has improved our understanding of the gaps between nutrition-related knowledge, attitudes, and practices [15‐17], these studies have mainly focused on preoperative nutrition-related components, and/or have employed an atheoretical design. Given the complexity of the clinical environment, it is important to employ empirically-supported theories to understand staffs’ behaviour. As such, establishing a better theoretical understanding of doctors’ decision-making is required to inform development of targeted strategies to improve postoperative nutrition practices among patients who have undergone colorectal surgery.

The Theoretical Domains Framework (TDF) is a theoretically-based approach to behaviour change. It consists of 14 domains that explain human behaviour in complex clinical environments: behavioural regulation; optimism; beliefs about consequences; skills; social/professional role and identity; emotion; beliefs about capabilities; goals; memory, attention and decision processes; reinforcement; intentions; environmental context and resources; social influence; and knowledge [18]. Consequently, this framework has been utilised in previous work to detect barriers and enablers to evidence-based guideline use among clinicans, and consequently inform interventions to increase uptake of evidence into habitual practice [19]. As such, the aim of this study was to use the TDF to explore the factors influencing medical staffs’ decision-making regarding postoperative nutrition prescription among non-critically ill colorectal patients.

A total of 21 medical staff were interviewed, including nine colorectal surgical consultants, three surgical fellows, four surgical trainees (two registrars and two principle house officers), and five junior medical doctors (one junior house officer and four interns). Nine and twelve staff were recruited at site one (S1) and site two (S2), respectively. No staff members who were approached declined to participate. Participant demographics are outlined in Table 1 and the medical career structure outlining the description of each position is depicted in Additional file 2. Approximately three-quarters of participants were male (n = 16). Through our analysis, we identified three themes and various subthemes which appeared to influence medical staffs’ decision-making. These are depicted in Table 2 and described in detail below. Of note, the use of “many” or “most” throughout this section is equivalent to ~ 75% or more of the sample in question.

This study explored factors influencing surgeons’ decision-making regarding the prescription of nutrition following colorectal surgery. Individual beliefs, patient-related factors and the social influence of peers were key factors that appeared to strongly influence staffs’ decision-making. These findings may be used to inform the development of educational and behaviour change strategies to improve postoperative nutrition prescription practices.

Surgeons held varied views on when they would usually or ideally prescribe nutrition after colorectal surgery. While most consultants, fellows and surgical trainees described prescribing fluids early in the postoperative period, only half said they would usually or ideally prescribe solids within 24 h after surgery. Behaviour change theories propose that people’s beliefs about the likely outcomes of an innovation and the value attached to those outcomes heavily influences their likelihood of adoption [23]. Indeed, surgeons’ perceptions of the risk-benefit-ratio of EOF appeared to influence prescription preferences in this study. Surgeons who preferred early solids and fluids (i.e. EOF advocates) believed this practice was safe and thought the benefits, such as improved return of gut function and reduced postoperative complications, infections and length of stay [4, 5] outweighed potential risks. Alternatively, staff hesitant of EOF consistently reiterated the risks of EOF and did not appear to view the practice as particularly advantageous. Collectively, these findings suggest that in order for surgeons to prescribe both early liquids and solids after colorectal surgery, they need to believe there is a relative advantage to this practice and that it is indeed safe.

Differences in surgeons’ perceptions of EOF are likely explained by their training and past experiences; findings consistent with previous studies exploring factors predicting adoption of evidence-based practice [24, 25]. Many EOF advocates discussed how ERAS had been at the forefront of their training and some had been exposed to formalised protocols implemented in other hospitals. This appeared to increase their confidence and knowledge in prescribing early fluids and solids. Alternatively, most surgeons who were hesitant of EOF were trained under the old feeding paradigm, where resumption of nutrition was indicated only after recovery of bowel function; a belief which remains entrenched as evidenced by surgeons’ responses. Changing long-standing attitudes and behavior of clinicians can be difficult [26]. Previous work shows education can increase knowledge, however may not immediately improve practice [27]. Rather, changing behavior is a continuous process whereby evidence-based innovations may undergo a lengthy period of negotiation among adopters, in which their meaning is discussed, contested, and reframed [28]. Such deliberation can increase or decrease the potential adopter’s perceived relative advantage of the innovation [29]. Indeed, this appeared to be evident within the current study, when a surgeon spoke of a negative EOF experience that occurred at one of the sites which decreased hesitant staff members’ confidence in prescribing early solids after colorectal surgery. For this reason, it is important to provide education and training on the benefits and safety of EOF, while also providing reassurance (e.g. through regular monitoring and feedback to staff) in the early stages of adoption to facilitate changes in surgeons’ prescription practices.

Behaviour change theories [23] recognise the role of the social environment in influencing behaviour. Individuals’ decision-making were directly and indirectly influenced by their seniors in the current study; a finding consistent with previous work [30]. The inter-professional influence of the attending consultant was evident, as fellows and surgical trainees expressed they prescribed diets in line with the consultant’s preferences, despite holding alternative views. For example, while most surgeons (particularly fellows and surgical trainees) specified their preference for using free fluids directly after surgery, clear fluids were still habitually prescribed at both hospitals as some consultants held this preference. The idea that doctors are expected not to question decisions of more senior staff has been identified in previous work [31]. These findings suggest that for EOF practices to align with evidence-based guidelines, behavioural change strategies particularly need to target consultants and fellows, given their direct influence on other staff. Further, changing behaviours and attitudes of senior staff towards EOF and nutrition in general will likely influence attitudes and perceptions of junior staff, considering surgeons are trained in highly hierarchical structures where practitioners tend to avidly follow the lead of their seniors; a finding demonstrated in the current study and previous work [32].

Lastly, patient-related factors, such as procedure type, demographics and clinical status appeared to strongly influence clinicians’ decision-making. Many surgeons described delaying the prescription of solids among patients who underwent procedures involving the small bowel or ascending colon due to “swelling” and “increased risk of ileus”. Indeed, evidence suggests there is a higher incidence of paralytic ileus after right-sided ileo-colic anastomosis compared with left-sided colo-colic and colo-rectal anastomosis [33, 34]. Further, in a recent study the safety of patients who underwent an ileostomy closure to have a full diet on POD1 were reported. However, this approach did result in a high incidence of vomiting (45%), which the authors hypothesised was due to oedema and spasm at the site of the anastomosis [35]. Hence, feeding recommendations may require adaptation for procedures involving the small bowel [35], particularly among elderly patients [33], who appear to be commonly considered as high-risk by surgeons. However, not all patient-related factors that influenced surgeons’ decision-making were evidence-based. For example, some staff spoke of using clear fluids over free fluids, soft or low fibre diets over full diet, and progressing patients’ diet types based on their bowel function; practices that are scientifically unfounded. Lastly, while many surgeons spoke of making nutrition care decisions in the “patient’s best interest”, this view was often held from their perspective and not in consultation with the patient. In fact, only a minority of surgeons described factoring patients’ food preferences and nausea or hunger ratings into their decision-making, while several stated they were better positioned to determine patients’ dietary needs. However, when patients participate in their care they have better outcomes and are more satisfied with their care [36, 37]. Given the relationship between shared decision-making and evidence-based practice is becoming increasingly recognised, strategies to support clinicians in enacting a more patient-centred approach are required.

While this study has provided valuable insight into factors influencing surgeons’ decision-making around postoperative feeding, it has several limitations. It is possible that some views were not represented in our sample, however we used purposive sampling and conducted interviews across two sites to improve generalisability and continued data collection until saturation was reached, which may increase the relevance of our findings for similar settings.

In summary, individual beliefs, patient-related factors and social influence of peers (particularly seniors) appeared to strongly influence surgeons’ decision-making regrading dietary prescription after colorectal surgery. As such, a multi-faceted approach to behavior change is required that targets each of these areas and is specific to the clinical context.