Outcomes following the introduction of an interdisciplinary shared decision-making clinic for older patients with colorectal cancer

Colorectal cancer (CRC) is the third most common cancer world-wide and the incidence is highest in older patients.¹ The standard of care for CRC includes surgical resection for stage 1–3 and selected cases of stage 4 disease, but older patients with CRC have high rates of concurrent multimorbidity² and frailty,³ which are known to increase perioperative risk.⁴ Pre-operative frailty has been associated with poorer outcomes after surgery for CRC, including prolonged length of hospital stay (LOS) and increased rates of major complications.⁵ Similarly, while surgery for all patients with CRC carries risk of morbidity and mortality,⁶ this risk is significantly increased in older patients who have increased LOS and rates of complications.⁷

Peri-operative risk must be weighed against the natural history of the disease in the context of patient life expectancy, quality of life and goals of care. Surgeons require an excellent understanding of the natural history of surgical disease and peri-operative risks but surgical decision making for older patients has become more complex due to increased life expectancy, multimorbidity and frailty. Interdisciplinary pathways for high-risk patients considering surgical treatment have been described to address this problem.⁸ In 2019, a shared decision-making (SDM) pathway for older patients with CRC being considered for surgical resection was instituted in Waikato Hospital, a tertiary referral centre in New Zealand. SDM pathway goals are to facilitate SDM for older patients with CRC who are moderate or high peri-operative risk and to facilitate peri-operative optimisation for patients proceeding to surgery.

Despite an increasing volume of literature supporting SDM in surgical patients,^8–10 there is no published literature reporting a dedicated SDM pathway for older patients with CRC. This study aimed to compare real-world outcomes for older patients with CRC considered for elective surgery before and after the introduction of an interdisciplinary SDM pathway at Waikato Hospital. The primary outcome was days alive out of hospital (DAOH). Secondary outcomes included mortality, treatment delivered and treatment deviation from colorectal multi-disciplinary meeting (MDM) recommendation (“gold standard” treatment recommendation).

Method

The standard and SDM clinical care pathways during the study period are shown in Figure 1. SDM consultation involves a holistic assessment in a single outpatient clinic visit, incorporating colorectal surgeon consultation, anaesthetic assessment and/or comprehensive geriatric assessment (CGA) by a specialist physician. Key stakeholders (patients, family/whānau and medical specialists) then participate in a values-based discussion of all treatment options and predicted risks and benefits drawing on the surgical, anaesthetic and geriatric assessments.

From 1 January to 31 December 2020, all patients diagnosed with CRC over 70 years of age discussed at the MDM were entered into a prospective database. Patients with CRC over 70 years of age discussed in the colorectal MDM from 1 January to 31 December 2018 were identified after screening all MDM lists. All patients with CRC in Waikato Hospital are discussed in the MDM prior to elective treatment, and usually prior to first specialist appointment (FSA) with a colorectal surgeon. Patient selection for the SDM pathway was determined by consultant medical specialists after screening patients over the age of 70 for multimorbidity or frailty using the Geriatric 8 (G8) score (score <14 used as indicator of frailty). Inclusion criteria for this study were age 70 years or over, discussed in CRC MDM, treated at Waikato Hospital in 2018 or 2020, being considered for surgical resection (palliative or curative intent) and confirmed histological diagnosis of adenocarcinoma of colon, rectum or appendix. Patients recommended for non-operative management by MDM or treated in the private sector were excluded.

Data were collected retrospectively from clinical records, including demographics, disease factors such as staging, tumour location and pathology results, assessment undertaken including specialist consultation, frailty scores and cognitive function, treatment factors including surgery vs non-operative care, deviation from MDM recommendation and outcomes including DAOH, mortality and LOS. DAOH was defined as the number of days the patient was alive excluding any days admitted to hospital (including all-cause emergency and elective admissions) in the 12 months from the day of surgery. Univariate statistical analysis used Chi-squared or Mann–Whitney U test as appropriate. Survival analysis was performed using the Kaplan–Meier method and the Mantel–Cox test. A two-sided alpha of 0.05 was considered statistically significant. Outcome analysis was performed of outcomes for patients who did not proceed to surgical treatment for CRC after SDM discussion.

Full ethics approval was obtained from the New Zealand Health and Disability Ethics Committee (HDEC 2022 FULL 12313).

View Figure 1–2, Table 1–4.

Results

One hundred and sixty-nine patients were included in the study. One hundred and three (60.9%) were male; the median age was 79 years (range 70–95) and was similar across both years of the study. Eighty-two out of 169 (48.5%) patients were treated in 2018 and 87/169 (51.5%) were treated in 2020. All patients had at least 12 months of follow-up performed. Twenty-seven out of 169 (16.0%) were diagnosed with stage 4 disease at time of discussion in MDM. Patient demographics and selected disease characteristics are presented in Table 1.

In 2018, no patients underwent formal frailty scoring or objective assessment of cognitive impairment. In 2020, 24 patients had a CGA performed, and 25 patients were seen formally in the SDM clinic with either anaesthetic or geriatric assessment performed. Frailty as a binary outcome of geriatric physician assessment, cognitive impairment scores and G8 scores are presented in Table 2.

Table 3 presents the treatment plans recommended by the MDM and treatment delivered for all patients in the study, as well as outcome data. After the introduction of the SDM clinic, a greater proportion of patients underwent non-operative management (16.1% vs 4.9%, p=0.021) and experienced a deviation in management from the MDM recommendation (18.4% vs 4.9%, p=0.008). DAOH was 6 days higher in 2020 after the introduction of the SDM clinic. There was no difference in LOS for surgical patients.

There was no statistically significant difference in all-cause mortality (Table 4) or survival (Figure 2) by Kaplan–Meier method using the Mantel–Cox test (2018 mean 45.5 months vs 2020 mean 30.9 months, median was not reached for either year, p=0.92) after introduction of the SDM pathway. Overall median survival was 34 months for those who did not have surgery; median was not reached for surgery group. For patients seen in the SDM clinic compared with those not seen in SDM clinic in 2020, there was no significant difference in overall survival (mean 33.2 months non-SDM and 30.0 months SDM, p=0.12). Median survival was not reached in either group. For the 25 patients seen in the SDM clinic, there was no difference in survival between patients who did and did not have surgery (mean 27.5 months without surgery, mean 31.2 months with surgery, p=0.22).

Twenty-five patients were formally seen in the SDM clinic in 2020. Fifteen out of 25 had elective surgery, 10/25 had planned non-operative management, of which one patient later underwent acute surgery, leaving nine patients managed non-operatively after SDM discussion. Two out of nine patients pursuing non-operative management died within 12 months of first MDM discussion and a further five patients had died at last follow-up. Causes of death were CRC-related (n=2), cerebrovascular accident (n=2), respiratory failure (n=1), cardiac failure (n=1) and unknown cause (n=1). One patient with significant multimorbidity and frailty who initially elected for non-operative management underwent emergency open right hemicolectomy for a bowel obstruction 10 months after MDM and died in hospital of heart failure 7 days post-operatively (355 DAOH from initial MDM).

Of the 15 patients who proceeded to elective surgery after the SDM clinic, seven patients had a right-sided colonic tumour, six patients had a left-sided colonic tumour, one patient had a rectal tumour and one patient had metastatic recurrence in the anterior abdominal wall. The median LOS was 8.9 days, median DAOH was 356 days (range 258–361 days). One patient in this group died within 12 months after undergoing a laparoscopic low anterior resection and loop ileostomy formation for rectal cancer followed by an 8-day post-operative hospital stay. Five unplanned re-admissions for high ileostomy output and cardiac disease resulted in an additional 16 days in hospital (333 DAOH), before the patient died of metastatic rectal cancer at home with palliative care support 357 days after the initial elective operation.

Discussion

This study demonstrates outcomes for older patients with CRC being considered for elective surgical management in a tertiary referral hospital in New Zealand before and after the introduction of an interdisciplinary SDM clinic. The results demonstrate a significant increase in non-operative management and deviation from MDM recommendation following introduction of the SDM clinic. This is the first report in published literature of an interdisciplinary SDM clinic specifically for older patients with CRC.

DAOH is a composite, patient-centred outcome measure describing a patient’s journey from a point in time, encompassing hospitalisation and mortality. DAOH has been gaining traction in surgical literature,^11,12 and was chosen in this instance because of implications for mortality, quality of life impact and resource consumption. It can also be calculated retrospectively from routinely collected data in the New Zealand health system. Median DAOH was 6 days higher after introduction of the SDM clinic; this difference may have resulted from selecting fewer patients for surgery with multimorbidity and frailty, which are likely to increase all-cause hospital readmission rates and mortality. This is a clinically important difference; considering the absolute mortality rate is low, this indicates a significant reduction in hospital bed days for patients undergoing surgery after SDM pathway introduction, although an accurate cost analysis is beyond the scope of this study.

There is evidence that consideration of high-risk surgery in older adults may not follow SDM principles when consultation is with a surgeon alone, as surgeons focus on risk assessment and informing patients of risks rather than taking part in a values-based SDM process.¹³ The addition of interdisciplinary specialists has potential to mitigate this. While a resource-intensive intervention, and while the current study does not include a cost–benefit analysis, there are significant potential cost savings from the SDM approach if major surgery is avoided in high-risk candidates at increased risk of complications and prolonged LOS. Rehabilitation needs can also be anticipated in advance, streamlining the post-operative pathway. Such a patient-centred, integrated approach also encompasses many aspects of tikanga Māori¹⁴ and therefore may be more acceptable to Māori patients, who are known to suffer from numerous ethnicity-based health inequities.¹⁵

There was no statistically significant difference in post-operative mortality or survival on Kaplan–Meier analysis after the introduction of the SDM clinic, but the study was not expected to demonstrate a difference in this metric due to total participant numbers and mortality being a rare event. A significant reduction in high-risk patients undergoing operative management leads the authors to hypothesise that patient selection for surgery was improved, with poorer candidates electing not to have surgery after SDM discussion. A larger study with statistical power to detect a true difference may help elucidate this. The high mortality rate prior to introduction of the SDM pathway should be noted; this was a driver for the quality improvement process.

A relatively small number of patients (n=25) were formally seen by an anaesthetist and/or geriatrician after the introduction of the SDM clinic. Because not all patients over 70 are seen, patient selection is reliant on the professional opinion of clinicians involved. However, the authors hypothesise that the benefit of the SDM pathway extends beyond those patients formally assessed, as the presence of an anaesthetist and geriatrician in the outpatient clinic alongside colorectal surgeons allows for ad hoc informal consultation. This discussion may result in formal SDM assessment, referral to another specialist and/or additional investigations. The senior clinicians involved have an excellent working relationship with free flow of information between them, and this working relationship itself may contribute to improved outcomes.

To maximise efficiency in a resource-constrained environment, MDM discussion for patients with new diagnoses of CRC in Waikato has historically occurred prior to colorectal FSA and therefore patient factors must be carefully considered when determining the final management plan in-clinic. Some patients must be re-discussed at the MDM if the optimal management pathway does not align with patient fitness and goals of care. It is likely that, with extensive values-based discussion and expert interdisciplinary consensus, good decisions are made in the SDM pathway, but this is reliant on the professional expertise of clinicians involved. Outcome assessment for patients managed non-operatively after SDM discussion found only 2/9 died of CRC-related causes. This suggests performing elective surgery would not have prolonged survival for at least 7/9 of these patients, irrespective of the occurrence of peri-operative complications, for which they would undoubtedly be a high-risk group. This provides some reassurance that patients are unlikely to be harmed by the shift to increased non-operative management.

There are important limitations of the study reported, and the findings associated with SDM pathway introduction cannot be claimed to be causative based on this retrospective, uncontrolled study. The implementation of the SDM clinic was driven by a desire to improve outcomes for older patients with CRC, and changes in care during the study period may have confounded the results. A new colorectal surgeon was employed in 2020, joining the four surgeons present in 2018. The average age of frailty onset has been reported as 69 years,¹⁶ but younger patients with frailty and multimorbidity have been excluded from this study. There are other important measures of success, including patient-reported outcomes such as decision regret and patient satisfaction with the SDM process. Further studies must explore these areas. A randomised study would be practically challenging, and to deny patients access to an established SDM clinic could be seen as unethical, as there are no apparent objective detrimental effects for the patient in a patient-centred SDM approach.

Conclusions

An SDM approach to older patients with CRC being considered for surgical resection combining colorectal surgeons, anaesthetists and geriatricians in a single consultation is a feasible option in a tertiary referral centre in New Zealand. The introduction of an SDM clinic in Waikato was associated with increased rates of deviation from MDM recommendation towards non-operative management and increased DAOH in patients undergoing surgery. An interdisciplinary SDM approach may improve patient selection for surgical resection and should be considered by those treating high-risk older patients with CRC. Future work should focus on patient-reported outcomes for patients seen in the SDM clinic.