Multidisciplinary Clinics for Colorectal Cancer May Not Provide More Efficient Coordination of Care

Abstract

Background

This retrospective study compares a multidisciplinary clinic (MDC) to standard care for time to treatment of colorectal cancer.

Methods

We queried our institutional ACS-NSQIP database for patients undergoing surgery for colorectal cancer from 2017–2020. Patients were stratified by initial clinic visit (MDC vs control). Primary endpoint was the time to start treatment (TST), either neoadjuvant therapy or surgery, from the date of diagnosis by colonoscopy.

Results

A total of 405 patients were evaluated (115 MDC, 290 Control). TST from diagnosis was not significantly shorter for the MDC cohort (MDC 30 days, Control 37 days; p=0.07) even when stratified by type of initial treatment of neoadjuvant therapy (MDC 30, Control 34 days; p=0.28) or surgery (MDC 32.5 days, Control 38 days; p=0.35).

Conclusion

Implementation of an MDC provides insignificant reduction in delay to start treatment for colorectal cancer patients as compared to standard care colorectal surgery clinics.

Introduction:

Colorectal cancer is the second most common cause of cancer related death in the United States with 104,000 new colon and 43,000 rectal cancer cases projected for 2020 [1]. The management options of colorectal cancer may include surgery, chemotherapy and radiation therapy in either an adjuvant or neoadjuvant fashion. These complex treatment algorithms require timely and efficient coordination of care between pathology, radiology, medical and radiation oncology, and surgical specialties [2,3]. While great strides have been made in the treatment of rectal cancer worldwide, there continues to be substantial variation in outcomes [4]. In order to standardize rectal cancer care within the United States, the National Accreditation Program for Rectal Cancer (NAPRC) was established through a collaboration between Optimizing the Surgical Treatment of Rectal Cancer (OSTRiCh) Consortium and American College of Surgeons. Accreditation requires demonstration of full compliance with set standards including a multidisciplinary team (MDT) approach to care [5]. The MDT approach has proven successful in improved clinical decision making, outcomes, and the patient experience in several cancer types including colorectal [6–9].

The multidisciplinary clinic (MDC) model was established as one way to facilitate an MDT approach by providing patients with the convenience of accessing timely, cohesive, integrated care by multiple specialists in one outpatient visit as opposed to the traditional model where interaction with multiple specialties occurs in temporally and geographically discrete encounters. The formation of these specialized clinics allows for easier, more direct communication amongst providers, more standardized, evidence-based management, and improved patient satisfaction and outcomes [10,11]. These multidisciplinary clinics require a significant amount of organizational, personnel and operational resources [11] but data on the effectiveness of these clinics in colorectal cancer, specifically, are sparse [12].

The objective of this study is to determine whether an MDC at a tertiary referral center improves timeliness of the treatment of colorectal cancer as compared to the standard care model. We hypothesized that MDC reduces the time from diagnosis to first clinic visit, complete staging, and initial treatment (surgery or neoadjuvant therapy). We further hypothesized that those benefits would be more profound for vulnerable patients such as minorities, the uninsured, or those travelling long distances to receive care.

Materials and Methods:

This retrospective study was approved by the University of Alabama at Birmingham (UAB) institutional review board. Patients undergoing colorectal surgery for colon or rectal cancer (ICD10: C18-C20) between January 2017 and February 2020 were selected using our institutional ACS National Surgical Quality Improvement (NSQIP) database. The patient medical charts were reviewed to ensure diagnosis and clinical stage. Subjects were excluded from this study if their diagnosis was anything other than colon or rectal adenocarcinoma, if they sought initial therapy at another institution, had recurrent cancer, or underwent an emergent procedure as initial therapy.

Clinical characteristics, sociodemographic factors, insurance status, and timelines for diagnosis and treatment were retrospectively reviewed. Overall stage (0-IV) was determined based on pathologic tumor/node/metastasis (TNM) classification and staging system. Patients were assigned to the control group if initially seen by radiation oncology, medical oncology, surgery or any other specialty in their respective clinics. Patients were included in the MDC group if initially seen by one or more specialties (medical oncology, surgical oncology, or radiation oncology) in the established multidisciplinary clinic. All unattached rectal cancer or complex colon cancer patients are scheduled in our multidisciplinary colorectal cancer clinic. Direct referrals to specific providers are scheduled in their own clinics. Insurance status was stratified into charity, self-pay, private, or public insurance. Distance was recorded as miles between patient’s home zip code and UAB hospital.

Start treatment date was defined as either initiation of chemotherapy, administration of the first radiation dose, or surgery. Initial surgery for symptom control such as diverting stoma prior to definitive surgery was included. The date of diagnosis was defined as date of colonoscopy. Date of complete staging for rectal cancer was defined as date of MRI or rectal ultrasound. Initial clinic visit was set as day of first consultation for the MDC group or first specialty appointment for control group. End of neoadjuvant therapy was logged as the day of final chemotherapy or radiation dose as recorded in clinic notes.

Primary outcome was the time to start treatment (TST) calculated in median days between day of diagnosis and beginning of initial treatment. Secondary outcomes were time intervals between other important phases of care including initial clinic visit, complete staging, and end of neoadjuvant therapy. Patients were then stratified by type of cancer (rectal vs colon), ethnicity and insurance types for further comparisons of TST from diagnosis between clinic types.

Categorical variables were described as percentages. Continuous data was described as medians and interquartile range. Differences between groups were tested using Wilcoxon and Kruskal Wallis tests as appropriate. TST from diagnosis was also modeled using distance to UAB as a predictor and plotted on a graph by clinic type. Patients with missing data were excluded from analysis for that particular piece. All analysis was performed at an alpha level of 0.05 using SAS 9.4 (SAS Institute, Cary, NC).

Results:

Patient Characteristics

Of 405 individuals who underwent surgery for colorectal adenocarcinoma and were eligible for this study, 290 were stratified to control group and 115 to MDC. Median age of all patients at time of surgery was 62 (26–89). The majority were male (52.4%) and Caucasian (71.6%) with private insurance (46.2%). Overall, most patients were seen for colon cancer (62.5%) and underwent surgery (61.5%) as initial treatment. Patients with rectal cancer had higher odds of receiving care at MDC compared to patients with colon cancer (OR 2.12; p<0.01). Similarly, patients whose initial treatment consisted of neoadjuvant therapy had higher odds of receiving care under the MDC model as compared to surgery only (OR 1.9; p=0.004). The clinics were different in terms of distribution of staging (p=0.006) in that patients with stage 0 (p=0.008) and stage 2 (p=0.003) cancers were more often seen in control clinics than in MDC. Median distance travelled did not differ between control and MDC (p=0.21). Further comparison of baseline characteristics of patients between control vs MDC groups are summarized in Table 1.

Table 1.

Baseline Characteristics

Baseline Characteristics	Total (N=405)	Control (N=290)	MDC (N=115)	P value
Age, median (range)	62 (26–89)	62 (26–89)	62 (37–83)	0.88
Male (%)	212 (52.3)	147 (50.7)	65 (56.5)	0.29
Ethnicity (%)				0.74
Caucasian	290 (71.6)	209 (72.1)	81 (70.4)
Black	107 (26.4)	75 (25.9)	32 (27.8)
Hispanic	3 (0.7)	1 (0.34)	2 (1.7)
Asian	5 (1.2)	5 (1.7)	0
Diagnosis (%)				<0.01*
Rectal Cancer	152 (37.5)	94 (32.4)	58 (50.4)
Colon Cancer	253 (62.5)	196 (67.6)	57 (49.6)
Stage (%)				0.006*
0	38 (9.4)	35 (12.1)	3 (2.6)
I	51 (12.6)	41 (14.1)	10 (8.7)
II	100 (24.7)	64 (22.1)	36 (31.3)
III	107 (26.4)	74 (25.5)	33 (28.7)
IV	31 (7.7)	25 (8.6)	6 (5.2)
Type of Start Treatment (%)				0.004*
Neoadjuvant	156 (39.0)	99 (34.1)	57 (49.6)
Surgery	249 (61.5)	191 (65.9)	58 (50.4)
Insurance Type (%)				0.21
Self-Pay/Charity	34 (8.4)	27 (9.3)	7 (6.1)
Private/Medicare/Medicaid	324 (80)	245 (84.5)	79 (68.7)
Unknown	47 (11.6)	18 (6.2)	29 (25.2)
Distance (mi), median (Q1, Q3)	67 (18–117)	65 (19–99)	75 (14–167)	0.21

Time intervals for control vs MDC

Time intervals between diagnosis and complete staging (MDC 13 days, Control 14 days; p=0.48), diagnosis and initial clinic visit (MDC 14.5 days, Control 18 days; p=0.10), and end of neoadjuvant therapy and surgery (MDC 63 days, Control 64 days; p=0.60) did not significantly differ between the two groups (Table 2). Time to start treatment (TST) from diagnosis was shorter for the MDC cohort but not significantly so (MDC 30 days, Control 37 days; p=0.07) even when stratified by type of initial treatment of neoadjuvant therapy (MDC 30, Control 34 days; p=0.28) or surgery (MDC 32.5 days, Control 38 days; p=0.35) (Table 2, Figure 1). Similarly, interval between first clinic visit and TST was not significantly different (MDC 14.5 days, Control 16; p=0.37), whether initial treatment was neoadjuvant therapy (MDC 14, Control 17; p=0.34) or surgery (MDC 15, Control 15; p=0.77).

Table 2.

Time intervals according to type of cancer

Overall	Total (N=405	Control (N=290)	MDC (N=115)	P value
Diagnosis to Complete Staging	13	14	13	0.48
Diagnosis to Initial Clinic Visit	17	18	14.5	0.10
End of Neoadjuvant Therapy to Surgery	64	64	63	0.60
TST	34	37	30	0.07
Diagnosis to Neoadjuvant Therapy	33.5	34	30	0.07
Diagnosis to Surgery	36	38	32.5	0.35
Rectal Cancer	Total (N=152)	Control (N=94)	MDC (N=58)	P value
Diagnosis to Complete Staging	13	14	13	0.48
Diagnosis to Initial Clinic Visit	20	21	16	0.09*
End of Neoadjuvant Therapy to Surgery	64	64	63	0.72
TST	32. 5	34	29	0.13
Diagnosis to Neoadjuvant Therapy	16	16.5	14	0.51
Diagnosis to Surgery	15	15	9	0.29
Colon Cancer	Total (N=253)	Control (N=196)	MDC (N=57)	P value
Diagnosis to Complete Staging	·	·	·	·
Diagnosis to Initial Clinic Visit	15	17	14	0.29
End of Neoadjuvant Therapy to Surgery	63	57	66	0.54
TST	36	39.5	32	0.26
Diagnosis to Neoadjuvant Therapy	33	39	27.5	0.10
Diagnosis to Surgery	36	39	32.5	0.37

Figure 1.

Time to start treatment (TST) from diagnosis according to first treatment modality

Time intervals for control vs MDC by cancer type

Time from diagnosis to complete staging (MDC 13 days, Control 14 days; p=0.48), to start therapy (MDC 29 days, Control 34 days; p=0.13), to neoadjuvant therapy (MDC 14, Control 16.5 days; p=0.51), and to surgery (MDC 9 days, Control 15 days; p=0.29), and time from end of neoadjuvant therapy and surgery (MDC 63 days, Control 64 days; p=0.72) did not significantly differ between clinic types when examining just the rectal cancer cohort (Table 2). Similarly, no difference was seen in time intervals between clinic types for colon cancer patients. The only difference was seen in time from diagnosis to initial clinic visit for rectal cancer patients (MDC 16 days, Control 21 days; p=0.09),

Time intervals between diagnosis and TST by social determinants of health

TST from diagnosis for controls with different insurance types (Medicare/Medicaid 39 days, Private 34 days, Self-pay 38 days, Charity 48.5 days; p=0.81) and ethnicities (Caucasian 35 days, Black 42 days, Hispanic 11 days, Asian 31 days; p=0.42), was similar. TST from diagnosis for MDC patients with different insurance types (Medicare/Medicaid 32 days, Private 30 days, Self-pay 23 days; p=0.51) and ethnicities (Caucasian 32 days, Black 32 days, Hispanic 14 days; p=0.12) was also not significantly different. TST from diagnosis was similar regardless of clinic for Caucasians (MDC 32 days, Control 35 days; p=0.17), minorities (MDC 28 days, Control 41 days; p=0.19), those with any type of private or public insurance (MDC 30 days, Control 37 days; p=0.12), and those who were self-pay or charity (MDC 23 days, Control 38 days; p=0.17) (Figure 2).

Figure 2.

Time to start treatment (TST) from diagnosis according to ethnicity and insurance status

Impact of distance travelled to UAB on TST from diagnosis was also analyzed. As distance to UAB increases, the probability that TST from diagnosis will be longer also increases to a significant degree in control patients (p=0.02). This positive relationship between distance and TST, however, was not statistically significant in MDC patients (p=0.5) (Figure 3).

Figure 3.

Predicted probability of time to start treatment (TST) as a function of distance to the hospital. TST from diagnosis was modeled using distance to UAB as a predictor that was then plotted on the graph by the clinic type. As the distance to UAB increases, the probability that it will take longer to go from diagnosis to start of therapy also increases. This positive relationship was significant for controls (p=0.02) but not for MDC (p=0.5).

Discussion:

Care provided by multidisciplinary teams as recommended by the NAPRC has undoubted benefits in colorectal cancer management and outcomes [2,6]. This study evaluates a colorectal cancer MDC as one way to facilitate the MDT approach at a tertiary referral center. We show that time intervals between various points in cancer care management for an MDC were not significantly shortened than for controls where patient interactions with specialists occurred in multiple single-provider clinics. Further, there was no baseline disparity in time to start treatment for either control or MDC due to ethnicity or insurance status, and thus, no demonstrated benefit of one clinic over the other. The only advantages provided by MDC over control were seen in rectal cancer patients in time from diagnosis to initial clinic visit and in patients with longer distances to travel where the increase in distance to the hospital increased the probability of delay in care for controls but not for MDC.

In an era of limited healthcare resources, it is imperative to provide quality care in an efficient manner. Specialized MDC’s are common in management of patients with multiple types of cancers with demonstrated improvement in overall outcomes, including survival [13,14]. One of the challenges with the traditional model in complex oncologic cases is optimizing coordination of care between multiple specialties, often at separate locations. The creation of MDCs is one strategy to facilitate communication amongst multiple providers [10], shortening time necessary for patients to complete all consultations and begin therapy as determined by a team [15]. Whereas multidisciplinary clinics have been shown to improve quality and efficiency of care, data assessing multidisciplinary clinics for colorectal cancer, specifically, are limited.

In this study, we focused on impact of an MDC on colorectal cancer care management up to the point of surgery. While simply reducing delays to initiation of therapy by a few days may not significantly improve clinical outcomes, this optimization of the oncologic care process may enhance patient satisfaction. Receiving a cancer diagnosis can be devastating and timely start of treatment assuages anxiety and boosts satisfaction [16]. However, our study did not show significant difference in time intervals to treatment from initial diagnosis in our new colorectal cancer MDC. This is in concordance to a previous study at another MDC that showed no difference in median time from diagnosis to initiation of neoadjuvant therapy [17]. We did see a shorter time interval from diagnosis to initial consultation in rectal cancer patients seen in MDC, but this difference did not persist to start of treatment.

Kozak et al. were able to show that initiation of treatment started more quickly for their MDC colorectal cancer cohort than non-MDC patients [18]. However, they only showed significant difference in time intervals from initial consultation once patients were already in the institution’s system and not from time of diagnosis, which could occur in another health system. While only measuring aspects under an institutions’ control eliminates multiple unknown factors involved in the referral process, it is important to also consider the MDC’s ability to process referrals with efficiency as this wait time has significant impact on patient satisfaction [19]. Our MDC cohort did not experience improvement in time to treatment from either diagnosis or initial consultation as compared to non-MDC patients. Additionally, since therapy cannot be initiated without staging, related time delays are important to consider. Neither study was able to show a difference in time to complete staging from diagnosis between MDC and control cohorts.

Health disparities for people with significant socioeconomic barriers must also be addressed when optimizing quality of care. In cancer patients, such barriers result in presentation of more advanced disease and worse prognosis [20]. Our study is unique in that it examined the impact of an MDC versus traditional clinics on timeline in cancer management for patients of different ethnicities, insurance status, and distances needed to travel for care. We saw no baseline disparity in TST from diagnosis in either MDC or control clinics when patients were stratified by ethnicity and insurance status. Additionally, when comparing MDC to control, there was no significant difference in time to initial therapy for Caucasians, minorities, and those with or without insurance, alike. In summary, because there was no disparity in efficiency of colorectal cancer care management under the standard care model for patients of different ethnicities and insurance status, there was no benefit provided by MDC to patients according to those social determinants.

Access to specialized cancer care is especially limited in rural areas, forcing patients to travel long distances for treatment, resulting in significant financial burden and several undesirable clinical outcomes, including later stage of diagnosis and less timely receipt of therapy [21,22]. Regardless, research and strategies for addressing these gaps have been slow to materialize [22]. Our study shows that while more distant residential zip codes from our facility predicted greater time delays to treatment for patients seen in standard clinics, this was not the case for MDC patients. Thus, we remain hopeful that providing these patients with an opportunity to access multidisciplinary care in one consolidated outpatient visit could be beneficial.

This is an observational study with inherent limitations. Patients were not randomized to the two cohorts; Rectal cancer patients were more likely to be seen in MDC. This nonrandomization likely introduced unknown, confounding variables impacting our results. In addition, referral bias may have been introduced. However, most early-stage cancers were seen in the standard clinics vs more late-stage cancer requiring more complex care were seen in MDC. Further, a selection bias may have been introduced as only patients that eventually needed surgical treatment for cancer were included, meaning a large cohort of patients requiring only medical therapy were excluded. Given added challenges of scheduling and coordinating surgery, this may have skewed the results.

In summary, our study on a colorectal cancer multidisciplinary clinic shows mixed results on coordination of care from diagnosis to staging, first clinic visit, and initiation of treatment. This MDC model of facilitating the nationally recommended multidisciplinary team approach to colorectal cancer treatment may not be the optimal route to improving time to care in high-volume centers as they may already be well equipped to manage complex cancer patients. Since existing data on this clinic model in colorectal cancer care is limited, future studies comparing MDCs across multiple institutions may be helpful in assessing and optimizing patient care flow. In addition to the focus on timeliness of care, assessment of an MDC’s effect on patient satisfaction and understanding of their disease and treatment outcomes in colorectal cancer is also needed. Lastly, studies should compare multidisciplinary clinics to other methods of facilitating communication between various specialties, such as teleconference, and streamlining interaction with patients through patient centered medical homes or telemedicine.

Highlights.

• Rectal cancer treatment requires management by multidisciplinary teams (MDT).

• The multidisciplinary clinic (MDC) model is thought to facilitate access to MDT.

• Data comparing MDC to standard clinics in colorectal cancer is sparce.

• We demonstrated insignificant reduction in time to start treatment for MDC.

• Only benefit of MDC was seen in time to therapy for patients living further away.