This is the first review to systematically investigate paper-based MDT QATs and DCs. Although a brief overview has previously been provided by Soukup et al.35, this review addresses a gap in the literature by providing detailed descriptions with specific and impartial quality assessments for each tool. It enables teams to identify and compare tool characteristics and make informed decisions. These tools can be used to monitor performance in line with NCAT7 standards. Evidence to suggest tool benefit in MDT QI is described. It is, however, envisaged that identification of their shortcomings will be of more benefit, identifying areas for more specific research and aiding the development of other tools in future.
Most QATs focused on assessing aspects of physical meetings, such as case information, leadership, attendance and teamwork. Governance, infrastructure and logistical elements of the MDT process were less frequently addressed. There were options for team self-assessment as well as observation. All QATs used Likert scales to assess each domain, with corresponding descriptions of optimal to suboptimal practice. There were no objective outcome measures. As they were used in isolation, the limitations of Likert scales should be considered36. One DC (MDT-QuIC23) was identified.
Although testing was usually limited to certain properties of validity and reliability, methodological quality in tool design was generally adequate. The concept and development of each tool was evidence-based and addressed some, if not all, of NCAT7 MDT quality domains. Tools were considered acceptable and clinician feedback was positive. Additionally, their simple nature makes them cost-effective and easily introduced.
Importantly, a single study, using the MDT-QuIC23 as part of a ‘quality improvement bundle’, did demonstrate a positive real-world impact on MDT discussion34. These results are encouraging, but are far from definitive - especially given the study’s limitations and mixed methods intervention. The paucity of studies using these tools for QI is reflective of the fact that, to date, they have mainly been utilised in observational research as the measure of quality, rather than the stimulus. This is an important distinction and highlights a significant void in the literature. These tools reasonably claim to be a method of identifying areas for improvement, but so far there is little evidence to substantiate this claim. A considerable amount of further research is required to better investigate their efficacy in QI. Given the nature of MDT discussion, randomised controlled trials are unlikely to be feasible, but controlled studies with QAT/DC-specific exposures would be beneficial to better demonstrate their role in creating change rather than simply measuring it.
Significantly, what these studies did not address was the effect tools had on the quality of the treatment decision itself. Tool domains closely reflect NCAT7 standards and, as such, they are compared to those in this review. It is important to understand, however, that these guidelines focus very much on the MDT process, rather than on what constitutes quality in the actual discussions and their outcomes. This raises the question of what ‘quality’ these tools are assessing and guiding towards. Clearly, an effective process is desirable, but correct and reproducible decisions will always be the most important indicator of MDT value.
Specific interest in discussion quality itself is growing, with some evidence suggesting that performance in this area is not always optimal37–40. Discussions tend to be dominated by biomedical information and led by surgeons and other diagnosticians38,40. Nurse specialist and other allied health professional input is more likely to be marginalised, ignored or non-existent37,41. These traditional hierarchies are potentially damaging, as unequal contribution defeats the purpose of collective expertise and opinion. Lanceley et al.38 also demonstrated the human nature of MDT discussion, highlighting the influence of personal experience and ethics. The potential for bias and groupthink in team decision-making is well known42 and MDTs are not excluded from this.
These factors could be extremely damaging to the MDT model, based as it is on the principle is that collective experience and decision-making is superior to single clinician lead care. Survey data suggests that clinicians are widely in agreement that MDT discussion is beneficial, but high quality evidence to prove this beyond doubt remains elusive43. Equally, there is limited data to evidence whether survival is truly improved by MDT discussion44–45. In their systematic review, Lamb et al.37 showed that MDT discussion did alter treatment decisions, but studies generally failed to correlate these changes with actual improvement in patient outcomes. Given the potential problems of team decision-making, this lack of definitive evidence certainly challenges the steadfast authority of the MDT within cancer services, as well as questioning their economic cost. Indeed, one study found a single MDT could cost up to £10,050 every month46. Ultimately, the tools presented in this review do not adequately assess MDT discussion and decision quality specifically or sufficiently enough to fully address these concerns. Once again, further investigation is required and future research should focus on ways to reliably assess discussions themselves and investigate effects on patient outcomes, rather than process quality alone.
Finally, the ‘unknown quantity’ in MDT decisions is patient choice. Autonomy is central to ethical healthcare and the importance of shared decision-making is enshrined in ‘Good Medical Practice’47. Notably, 519,21–22,23−24 of the 6 tools identified in this review did incorporate scoring for (indirect) discussion of patient views. However, given their scope, this remained a small part of the overall assessment. Currently, patient involvement in MDT decisions does appear to be limited37. Tellingly, one study found only 4% of investigated MDTs directly involved patients in their own treatment discussion48. Interestingly, evidence suggests that nurses are more likely to advocate for patients in the decision-making process35,37, further reinforcing the importance of equal participation. In light of the apparent barriers to patient involvement, calls to review the process have been made49. Could the nurse specialist have a bigger role in the discussion by proactively presenting the patient’s views? Or should the MDT outcome be a range of options that are then presented to the patient in clinic? What is clear is that greater integration of psychosocial factors will only add to the complexity of treatment decisions, making consistency and structural solidity even more essential. Tools aiding standardisation in the process may therefore have a greater role in the MDT of the future.
As MDTs evolve, digital solutions are also likely be utilised more frequently. Of these, decision support systems11–12, 50 may be particularly advantageous, as they offer rapid integration of patient information with evidence-based guidelines to generate objective management options. Early evidence has shown that this can increase guideline compliance and appropriate trial recruitment11. Going forwards, a combination of tools and technologies could be used to achieve the goal of high standards and reproducibility in MDT decision-making.
In summary, this review identifies and presents several paper-based tools for assessing the MDT process and guiding team discussions. Methodological quality was generally acceptable. These tools were developed to measure against, and increase compliance with, accepted high standards in the general MDT process. They represent a practical and relatively simple intervention that teams could employ to monitor their performance according to those standards and potentially identify areas for improvement. Extremely limited and relatively poor quality evidence supports the use of one tool in facilitating elements of MDT QI. Whether these tools overall could have a positive impact on decision quality and, crucially, on patient outcomes has not been established.