Health Care Delivery
Original Contribution
Reasons for Delay in Time to Initiation of Adjuvant Chemotherapy for Colon Cancer By David W. Wasserman, MD, Majdi Boulos, MD, Wilma M. Hopman, MA, Christopher M. Booth, MD, Rachel Goodwin, MSc, MD, and James J. Biagi, MD
Abstract Purpose: Adjuvant chemotherapy (AC) improves survival among patients with colon cancer (CC). Two meta-analyses have demonstrated a decrease in survival with increasing time to AC (TTAC). Here, we examine the predominant factors leading to delay in TTAC.
Methods: Individual medical records of 580 patients with CC who initiated AC August 2005-November 2010 at two large academic cancer centers in Eastern Ontario were reviewed. Information regarding patient, disease, and treatment characteristics, including time intervals between each step in the cancer care pathway from surgery to AC, was captured. Patients were then categorized into three groups for comparison: (I) postoperative complication, (II) oncologist- or patient-initiated delay, (III) no delay. These groups were compared using 2 tests and one-way analysis of variance. A multivariable logistic regression model was used to determine factors associated with TTAC ⬎ 8 weeks in all patients and in group 1 alone.
Introduction In large phase III randomized clinical trials, adjuvant chemotherapy (AC) for colon cancer has been shown to significantly improve overall survival in stage III disease and trend toward increased disease-free survival in selected patients with stage II disease at high risk of relapse.1-4 In these studies, the first cycle of chemotherapy was administered within 8 weeks of surgery. However, in routine clinical practice, AC is often initiated more than 8 weeks after surgery.5 Two recent meta-analyses analyzing time to adjuvant chemotherapy (TTAC) in colon cancer have shown that the survival benefit of AC diminishes with increasing wait times after surgery.6-7 Our group has previously demonstrated that each 4-week increase in TTAC is associated with a significant decrease in both overall survival (hazard ratio [HR] ⫽ 1.14; 95% CI, 1.10 to 1.17) and disease-free survival (HR ⫽ 1.14; 95% CI, 1.10 to 1.18). Des Guetz et al7 used a dichotomous variable of 8 weeks and found that delaying AC for ⬎ 8 weeks was associated with worse overall survival (rate ratio ⫽ 1.20; 95% CI, 1.15 to 1.26). Few studies have examined the reasons underlying delays in TTAC. Here, we examine individual patients’ medical records from two large academic cancer centers in Ontario to determine reasons for delay that may not be captured with commonly used administrative data. We also analyze wait times along each step in the pathway of care from surgery to chemotherapy. We hypothesized that the majority of cases of TTAC ⬎ 8 weeks would e28
JOURNAL
OF
ONCOLOGY PRACTICE
•
Results: TTAC among the three groups was (I) 10.1 ⫾ 2.7 weeks, (II) 10.5 ⫾ 3.6 weeks, (III) 8.5 ⫾ 2.1 weeks (P ⬍ .001). The only significant predictor of TTAC ⬎ 8 weeks on multivariable analysis in group I was route of AC via central venous catheter (odds ratio [OR] ⫽ 2.4; 95% CI, 1.2 to 4.9). When multivariable analysis was performed on all patients, the presence of postoperative complications (OR ⫽ 2.4; 95% CI, 1.6 to 3.8) and oncologist- or patient-initiated delay were the strongest predictors of delay (OR ⫽ 3.5; 95% CI, 2.1 to 6.0). The percentages of patients with TTAC ⬎ 8 weeks were (I) 76.4% (n ⫽ 110), (II) 81.4% (n ⫽ 92), (III) 57.9% (n ⫽ 187). Conclusions: In patients with no reason for delay, most experienced TTAC ⬎ 8 weeks. This likely reflects delays in referral, consultation, and chemotherapy booking. These health-system factors are modifiable, and future quality improvement initiatives should focus on how to reduce them.
be due to health care system delays, specifically, the cumulative wait times between referrals, consultations, procedures, and finally chemotherapy. We also hypothesized that the minority of cases of TTAC ⬎ 8 weeks would be due to postoperative complications, physician-initiated delays, or patient-initiated delays.
Methods Study Population Ontario has the largest population of Canada’s provinces and territories. Universal health care is provided through a publically funded system, with a minor component of a parallel private system. Cancer care is overseen by Cancer Care Ontario and is administered through 14 local health integration networks (LHIN). Each LHIN has a regional comprehensive cancer center.8 There are six LHINs that contain academic medical centers. In these LHINs, the regional comprehensive cancer center is housed at the academic institution’s affiliated hospitals. We included two of the six academic regional cancer centers in Ontario for our study: The Ottawa Hospital Cancer Centre, with a catchment area of 1.6 million people,9 and the Cancer Centre of Southeastern Ontario in Kingston, Ontario, with a catchment area of 500,000 people.10 Appropriate research ethics board approval was obtained at both centers, and a datasharing agreement was completed. At both institutions, all chemotherapy is ordered through the software program OPIS
V O L . 11, I S S U E 1
Copyright © 2014 by American Society of Clinical Oncology
Information downloaded from jop.ascopubs.org and provided by at Charité - Med. Bibliothek on May 5, 2015 from 130.133.8.114 Copyright © 2015 American Society of Clinical Oncology. All rights reserved.
Queen’s University; Kingston General Hospital, Kingston; and University of Ottawa, Ottawa, Ontario, Canada
Delays in Time to Adjuvant Chemotherapy
2005. Using the OPIS 2005 chemotherapy administration database, patients with pathologically confirmed colon cancer who started AC between August 1, 2005 and November 30, 2010 were identified and had their medical records abstracted.
Data Abstraction
Copyright © 2014 by American Society of Clinical Oncology
Data were collected in a Microsoft Excel file and imported into IBM SPSS (version 20.0 for Windows) for statistical analyses. The time data were initially graphed to assess the underlying distribution. Data were then analyzed descriptively, including percentages and frequencies for categorical data (sex, comorbidities, surgical hospital, and stage); and means, standard deviations, medians, and quartiles for continuous data (time data, length of stay, and age). Analyses used the three aforementioned groups for comparison: Group I, Group II, and Group III. The groups were compared using 2 tests for categorical data, one-way analysis of variance for continuous data, and the Kruskal-Wallis test for length of stay. Factors associated with TTAC ⬎ 8 weeks were identified using 2 tests and independent samples t tests. Multivariable logistic regression was then used to assess which variables predict for TTAC ⬎ 8 weeks in all patients and in Group III group only. A two-sided P value ⬍ .05 was considered statistically significant.
Results Figure 1 shows the flow diagram of the 580 patients analyzed in our study. The number of patients categorized into the three groups was as follows: (I) 24.7% (n ⫽ 143), (II) 19.5% (n ⫽ 113), (III) 55.9% (n ⫽ 324). Therefore, three quarters (75.4%) of the patients had no medical or surgical reason for TTAC delay. The most common reason for delay in Group II was patient decision to delay treatment to consider whether he or she wanted to proceed with AC (n ⫽ 61, 54.0%). The most common postoperative complication in Group I was ileus (n ⫽ 49, 34.2%). Table 1 lists baseline patient characteristics and time data based on group categorization. As shown in Table 1 and Appendix Figure A1 (online only), total TTAC and the subintervals differed among the three groups. The only exception to this was the interval between surgeon referral and MO consultation, which was consistent among the three groups. Group III had the lowest TTAC, but the mean was still ⬎ 8 weeks (8.5 ⫾ 2.1 weeks). As would be expected, the time from MO consultation to AC was longest in Group II, and time from surgery to referral to MO was longest in Group I. Median length of stay in hospital after surgery in days was 10.0 (95% CI, 6.0 to 14.0), 6.0 (95% CI, 5.0 to 7.0), and 6.0 (95% CI, 4.0 to 7.0) for Group I, Group II, and Group III, respectively (P ⬍ .001, KruskalWallis). There were some notable differences in group characteristics. Oral chemotherapy was administered more frequently in Group II (37.2%), compared with Group I (23.8%) and Group III (17.9%; P ⫽ .001). There were more patients with comorbidity in Group I compared with Group II and Group III (50% v 39% and 38%, respectively; P ⫽ .054). There was no significant difference in mean age of the groups. Table 2 displays univariable and multivariable analysis of clinical, disease, and treatment factors associated with TTAC ⬎ 8 weeks using the 324 patients in Group III as well as all 580
J A N U A R Y 2015
•
jop.ascopubs.org
e29
Information downloaded from jop.ascopubs.org and provided by at Charité - Med. Bibliothek on May 5, 2015 from 130.133.8.114 Copyright © 2015 American Society of Clinical Oncology. All rights reserved.
All data abstraction was by performed by D.W.W. and M.B. in Kingston and Ottawa, respectively, coordinated with a test cohort and characterization by consensus to ensure consistency. The following data were collected from the patients’ electronic medical record (EMR): demographics, academic versus community surgical hospital, tumor stage, baseline prespecified comorbidities, length of stay in hospital after colon resection, postoperative complications or intercurrent medical illness, patients’ medical fitness for chemotherapy at time of medical oncology (MO) consultation, whether chemotherapy was ordered directly after initial MO consultation, why AC was not ordered directly after MO consultation (if applicable), route of chemotherapy administration, and time parameters. Time parameters collected included dates of surgery, referral to MO, MO consultation, central venous catheter (CVC) insertion (if applicable), and first cycle of chemotherapy administration. The MO consultation note served as the primary source of information for comorbidities and postoperative complications, with the discharge summary and EMR serving as secondary sources of information. Prespecified comorbidities for inclusion included cardiovascular, respiratory, renal, diabetes, and other. Patients were dichotomized into presence or absence of comorbidities, as well as presence or absence of postoperative complications. Patients’ fitness for chemotherapy at initial MO consultation was collected. Chemotherapy was determined to be ordered directly after initial consultation if the oncologist indicated that arrangements for chemotherapy would commence immediately. A patient was determined to be medically fit for chemotherapy unless explicitly stated otherwise in the consultation note. When chemotherapy was not ordered directly after the consult, the reason why it was not was determined from the consultation note. Subsequent clinic notes were examined if there was ambiguity about the reason from the consultation note. Patients were divided into three groups. Group I includes all patients who experienced a postoperative complication or intercurrent illness between surgery and MO consultation. Group II consists of patients with no postoperative complication or intercurrent medical illness, but in whom AC was deferred after MO consultation as a result of patient preference or the need for further investigations such as restaging. Group III includes patients without a postoperative complication or intercurrent illness between surgery and MO consultation who underwent AC directly after MO consultation without additional delay due to patient preference or need for further investigations. We considered splitting Group I into “fit for AC at time of MO consultation” and “not fit for AC at time of MO consultation” subgroups, but there would have been too few patients in the “not fit” subgroup for meaningful analysis.
Statistical Analyses
Wasserman et al
Initial search results (N = 640) Patients not eligible as a result of missing data (n = 60) Eligible patients (n = 580)
(Group II; n = 113) AC not ordered after consult Patient decision (n = 61) Restaging (n = 33) Evaluation for clinical trial (n = 13) Tumor board review (n = 8) Access to chemotherapy/approval (n = 6) Other (n = 15)
Postoperative complication or intercurrent medical illness (Group I) (n = 143)
(Group III; n = 324) AC ordered after consult
Medically fit for chemotherapy (n = 111) at time of MO consultation Ileus/SBO (n = 49) (n = 16) Wound infection Other infection (n = 14) Cardiovascular events (n = 10) Wound dehiscence (n = 10) VTE (n = 8) (n = 5) Pneumonia (n = 5) Anastomotic leak (n = 5) Postoperative bleeding Poor functional status (n = 5) leading to prolonged hospitalization/rehabilitation Other (n = 9)
Medically unfit for chemotherapy at time of MO consultation Wound healing/infection Nononcologic specialist follow-up Other infection ECOG > 2 Other
(n = 32) (n = 17) (n = 5) (n = 5) (n = 4) (n = 3)
Figure 1. Flow diagram of the patient population included in the study. AC, adjuvant chemotherapy; ECOG, Eastern Cooperative Oncology Group performance status; MO medical oncology; SBO, small bowel obstruction; VTE, venous thromboembolism.
patients. The only variable that was significantly associated with TTAC ⬎ 8 weeks in Group III was intravenous AC via CVC (odds ratio [OR] ⫽ 2.4; 95% CI, 1.2 to 4.9; P ⫽ .013). Notably, age and presence of comorbidities had no significant association with TTAC ⬎ 8 weeks. The strongest predictors of TTAC ⬎ 8 weeks in all 580 patients were the presence of a postoperative complication or intercurrent illness (OR ⫽ 2.4; 95% CI, 1.6 to 3.8; P ⱕ .001) and oncologist- or patientinitiated delay (OR ⫽ 3.5; 95% CI, 2.1 to 6.0; P ⱕ .001). There was no significant association for comorbidities, age, stage, or type of surgical hospital. Again, AC delivered IV with CVC was significantly associated with TTAC ⬎ 8 weeks compared with oral or IV without CVC (OR ⫽ 1.7; 95% CI, 1.0 to 2.9; P ⫽ .037). The percentages of patients with TTAC ⬎ 8 weeks in the three groups were (I) 76.9% (n ⫽ 110), (II) 81.4% (n ⫽ 92), and (III) 57.7% (n ⫽ 187). In other words, only 23.1% of Group I patients, 18.6% of Group II patients, and 42.3% of Group III patients received AC within 8 weeks. Importantly, there was no effect on TTAC if surgery was performed in an affiliated hospital versus a cancer center hospital (P ⫽ .32).
Discussion There are multiple available studies that have examined the association between delayed TTAC and survival in colorectal cancer. However, there is a paucity of data about the etiology of these delays. Two studies have examined administrative databases to answer this question.11,12 The first demonstrated that the presence of a postoperative complication or intercurrent e30
JOURNAL
OF
ONCOLOGY PRACTICE
•
illness was a significant predictor of increased TTAC with relation to rectal cancer adjuvant therapy in an American administrative database. They also observed that advanced age and black race were associated with greater delays to AC, but only in the group of patients without established contact with a medical oncologist before surgery. This paradigm of “established contact” is not applicable to colon cancer because medical oncologists do not see patients in consultation before surgery. We did not find any significant association between age and TTAC on multivariable analysis. Race was not examined as this information is not captured routinely in Canadian administrative health data sets. Others have previously suggested that comorbidity and postoperative course confound the relationship between delayed TTAC and decreased overall survival.13 They speculated that those who experience delayed TTAC are likely to have more comorbidity and are therefore more likely to have decreased survival. Our results demonstrate that TTAC is delayed among patients with postoperative complications. However the most notable findings of our study are (1) the majority of patients (75.4%) do not have any medical reason to delay AC, and (2) most of these patients (63.8%) have TTAC delayed ⬎ 8 weeks. The second study to investigate causes of delay in TTAC demonstrated, through administrative datasets in the province of Alberta, Canada, that patients who received chemotherapy ⬎ 12 weeks postsurgery had more comorbidities compared with patients who received AC within 12 weeks of surgery.12 Had their analysis been dichotomized at 8 weeks, it is uncertain whether there would have been a difference. They did not have
V O L . 11, I S S U E 1
Copyright © 2014 by American Society of Clinical Oncology
Information downloaded from jop.ascopubs.org and provided by at Charité - Med. Bibliothek on May 5, 2015 from 130.133.8.114 Copyright © 2015 American Society of Clinical Oncology. All rights reserved.
No postoperative complication or medical illness (n = 437)
Delays in Time to Adjuvant Chemotherapy
Table 1. Baseline Characteristics of 580 Patients Treated With Adjuvant Chemotherapy for Resected Colon Cancer at the Cancer Centre of Southeastern Ontario and the Ottawa Hospital, 2005 to 2010 Group I (n ⴝ 143) Characteristic
Mean
Age, years (range 23-90) No.
Group II (n ⴝ 113)
Group III (n ⴝ 324)
SD
Mean
SD
Mean
SD
10.7
65.2
11.9
62.9
10.8
No.
%
%
No.
%
Sex 85
59.4
60
53.1
168
51.9
58
40.6
53
46.9
156
48.1
Cancer Centre No
64
44.8
52
46.0
159
49.1
Yes
79
55.2
61
54.0
165
50.9
T classification T3
79
59.0
66
64.1
185
64.0
T4
55
41.0
37
35.9
103
35.6
N classification N0
35
24.5
30
26.5
54
16.7
N1
63
44.1
58
51.3
160
49.5
N2
45
31.5
25
22.1
109
33.7
Comorbidities No
72
50.3
69
61.1
201
62.0
Yes
71
49.7
44
38.9
123
38.0
Chemotherapy Oral IV with CVC IV without CVC
34
23.8
42
37.2
58
17.9
100
69.9
62
54.9
246
75.9
9
6.3
9
8.0
20
6.2
Chemotherapy ordered after consult No
57
39.9
113
100.0
0
0.0
Yes
86
60.1
0
0.0
324
100.0
Mean
SD
Mean
SD
Mean
SD
Time data, weeks Surgery to MO referral
3.6
⫾1.9
2.7
2.0
3.1
1.8
Referral to MO consultation
3.4
1.9
3.5
2.4
3.3
1.6
Consult to CVC*
2.5
1.7
4.0
3.6
1.9
1.1
1.4
1.0
1.4
0.6
No. CVC to chemotherapy*
98 0.8
No. Consult to chemotherapy Total time to adjuvant chemotherapy
62
98
245
62
0.7 245
3.2
2.4
4.7
3.4
2.2
1.3
10.1
2.7
10.5
3.6
8.5
2.1
NOTE. Group I comprises patients who experienced a postoperative complication or intercurrent medical illness; Group II comprises patients without a postoperative complication and/or intercurrent illness but in whom chemotherapy was deferred at time of medical oncology consultation as a result of either patient preference and/or need for further investigations; Group III comprises patients in whom adjuvant chemotherapy was ordered directly after MO consultation without preceding postoperative complication and/or intercurrent illness. Abbreviations: CVC, central venous catheter; IV, intravenous; MO, medical oncology. * CVC line insertion date not known for two patients in Group I and one patient in Group III.
access to data regarding postoperative complications, but they did have access to extensive demographic data. They reported that patients with TTAC ⬎ 12 weeks were more likely to live in neighborhoods with a high percentage of divorced, separated, or widowed; a low employment rate; and a low median household income. On average, in all three groups, the time between having surgery and MO consultation was between 6 and 7 weeks. This Copyright © 2014 by American Society of Clinical Oncology
leaves little time for a patient to decide on chemotherapy, complete restaging if necessary, obtain a CVC, and then book chair time in the chemotherapy suite. In patients without a postoperative complication or intercurrent illness, much greater effort should be made to have assessment by medical oncology sooner than is done in current practice. If time from surgery to MO consultation or time from referral to MO consultation could be decreased by an average of 1 week, then the average TTAC
J A N U A R Y 2015
•
jop.ascopubs.org
e31
Information downloaded from jop.ascopubs.org and provided by at Charité - Med. Bibliothek on May 5, 2015 from 130.133.8.114 Copyright © 2015 American Society of Clinical Oncology. All rights reserved.
Male Female
Wasserman et al
Table 2. Variables Associated With Initiation of Adjuvant Chemotherapy More Than 8 Weeks After Surgery for Patients With Colon Cancer Treated at the Cancer Centre of Southeastern Ontario and the Ottawa Hospital, 2005-2010 Proportion > 8 wk Characteristic
No.
Univariable Analysis
%*
OR
95% CI
Multivariable Analysis P
OR
95% CI
P
Patients in Whom AC Was Ordered Directly After MO Consult (n ⫽ 324) Patient related Age, years 93
55.7
Ref
Ref
65-74 (n ⫽ 108)
70
64.8
1.47
0.89 to 2.42
.13
1.62
0.95 to 2.75
.077
ⱖ 75 (n ⫽ 49)
25
51.0
0.83
0.44 to 1.57
.56
1.05
0.50 to 2.22
.89
Male (n ⫽ 168)
98
58.3
Female (n ⫽ 156)
90
57.7
0.97
0.63 to 1.52
.91
1.11
0.70 to 1.75
.67
1.15
0.73 to 1.82
.54
1.21
0.73 to 1.98
.46
Sex Ref
Ref
Relevant medical comorbidity† Yes (n ⫽ 123)
74
60.2
No (n ⫽ 201)
114
56.7
Ref
Ref
Disease related Stage‡ 2 (n ⫽ 54)
28
51.9
3 (n ⫽ 269)
159
59.1
105
63.6
83
Ref
Ref
0.75
0.41 to 1.34
.33
1.07
0.55 to 2.08
52.2
0.62
155
63.0
9
45.0
24
41.4
.83
0.40 to 0.97
.038
0.64
0.40 to 1.02
.061
2.41
1.35 to 4.32
.003
2.43
1.21 to 4.90
.013
1.16
0.42 to 3.23
.78
0.98
0.32 to 2.95
.97
Treatment related Surgical hospital Cancer Centre hospital (n ⫽ 165) Affiliate (n ⫽ 159)
Ref
Ref
Chemotherapy IV with CVC (n ⫽ 246) IV without CVC (20) Oral (n ⫽ 58)
Ref
Ref Results for All 580 Patients
Patient related Age, years ⬍ 65 (n ⫽ 285)
185
64.9
65-74 (n ⫽ 192)
133
69.3
1.22
0.82 to 1.80
.32
1.30
0.86 to 1.99
.22
ⱖ 75 (n ⫽ 103)
71
68.9
1.20
0.74 to 1.94
.46
1.35
0.76 to 2.38
.31
Male (n ⫽ 313)
209
66.8
Female (n ⫽ 267)
180
67.4
1.03
0.73 to 1.46
.87
1.14
0.79 to 1.64
.48
1.01
0.71 to 1.44
.95
0.97
0.66 to 1.43
.88
Ref
Ref
Sex Ref
Ref
Relevant medical comorbidity Yes (n ⫽ 238)
160
67.2
No (n ⫽ 342)
229
67.0
Ref
Ref
Disease related Stage‡ 2 (n ⫽ 119)
82
68.9
3 (n ⫽ 460)
306
66.5
Cancer Centre hospital (n ⫽ 305)
212
69.5
Affiliate (n ⫽ 275)
177
64.4
Ref 1.12
0.72 to 1.72
0.79
0.56 to 1.12
Ref .62
0.89
0.55 to 1.44
.19
0.83
0.58 to 1.20
.63
Treatment related Surgical hospital Ref
Ref .32
Reason for delay Postoperative complication (n ⫽ 143)
109
76.2
2.35
1.51 to 3.67
⬍.001
2.45
1.56 to 3.85
⬍.001
Oncologist/patient initiated (n ⫽ 113)
92
81.4
3.19
1.89 to 5.38
⬍.001
3.52
2.05 to 6.03
⬍.001
188
58.0
No reason (n ⫽ 324)
Ref
Ref continued on next page
e32
JOURNAL
OF
ONCOLOGY PRACTICE
•
V O L . 11, I S S U E 1
Copyright © 2014 by American Society of Clinical Oncology
Information downloaded from jop.ascopubs.org and provided by at Charité - Med. Bibliothek on May 5, 2015 from 130.133.8.114 Copyright © 2015 American Society of Clinical Oncology. All rights reserved.
⬍ 65 (n ⫽ 167)
Delays in Time to Adjuvant Chemotherapy
Table 2. (Continued) Proportion > 8 wk Characteristic
No.
%*
Univariable Analysis OR
95% CI
Multivariable Analysis P
OR
95% CI
P
Chemotherapy IV with CVC (n ⫽ 408)
280
68.6
1.22
0.81 to 1.84
.34
1.73
1.03 to 2.88
.037
IV without CVC (38)
23
60.5
0.86
0.41 to 1.79
.68
0.97
0.44 to 2.15
.94
Oral (n ⫽ 134)
86
64.2
Ref
Ref
would drop below an 8-week goal in patients with no justifiable reason for delay. Another important finding is that most patients who requested a delay to contemplate AC ended up deciding on oral chemotherapy rather than IV. We can see this relationship in Table 1, whereby the highest proportion of patients undergoing oral chemotherapy is in the Group II group. The highest proportion undergoing AC with CVC is in Group III. This may represent the more motivated patient subgroups that eagerly want to pursue the most efficacious treatment. On the basis of our review of individual patient medical records, we were able to examine time intervals along the continuum of care from time of surgery to start of chemotherapy, which previous studies were unable to do. Furthermore, we could identify Group II, which would not be ascertained from an administrative database. In our analysis, the two strongest factors that predict delayed TTAC are the presence of a postoperative complication or intercurrent illness and oncologistor patient-initiated delay, but as hypothesized, the absolute number of patients with TTAC ⬎ 8 weeks was highest in Group III, reflecting health system delays. The results of our study have led to practice changes locally, such as computerized physician notification of pathology results for patients with colorectal cancer, enhanced nurse navigator roles, and timely central line booking. There are limitations to our study. The first is that it was performed retrospectively. Second, we did not have access to socioeconomic data to examine whether socioeconomic factors played a role in our population. Third, there may be other contributing variables that we did not capture in our analysis, such total dose of chemotherapy administered or baseline serum carcinoembryonic antigen level. Fourth, our index of comorbidity was based on conditions reported in the surgical discharge and oncology clinics notes. Accordingly, it is possibly that comorbidity may be under-reported in this study. Fifth, because our study population is from a single, albeit large, geographical region, clinicians should be attentive when generalizing our results to their patient population. Lastly, we did not include patients seen by MO who declined chemotherapy altogether. However, there is no reason to suggest that the time to Copyright © 2014 by American Society of Clinical Oncology
seeing a medical oncologist in those patients would differ significantly from those who accept AC. In summary, our study indicates that the majority of patients with resected colon cancer do not have a clinical reason to justify delays in TTAC. The presence of comorbidities was uncommon in our population and when present had no relationship with TTAC ⬎ 8 weeks on multivariable analysis. The presence of a postoperative complication or intercurrent illness and oncologist- or patient-initiated delay are significantly associated with delays in TTAC ⬎ 8 weeks. Postoperative complication or intercurrent illness are beyond the control of a medical oncologist and are not easily modifiable. Oncologist- or patient-initiated delay is often unavoidable when a patient is unsure of whether or not to proceed with chemotherapy and requires time for careful consideration. Furthermore, we believe that if there is concern regarding metastases, restaging investigations should not be omitted solely to decrease a patient’s TTAC. Fortunately, these situations occur in the minority of patients with resected colon cancer. We should therefore direct our attention to the population of patients without any clinically identifiable reason for delay, which represent the majority of patients, but of whom only half initiate chemotherapy within 8 weeks from surgery. Examination of the time intervals along the pathway of care shows that there is room for improvement in referral from surgery to MO, time from referral to MO consultation, and from MO consultation to CVC insertion. Future quality improvement initiatives should identify strategies to decrease these health system delays. Our results do not allow us to define an optimum TTAC, but practically an 8-week maximum for most patients may be reasonable. Acknowledgment Presented in poster format at the 10th Gastrointestinal Cancers Symposium, San Francisco, CA, January 24-26, 2013. Authors’ Disclosures of Potential Conflicts of Interest Although all authors completed the disclosure declaration, the following author(s) and/or an author’s immediate family member(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a “U” are
J A N U A R Y 2015
•
jop.ascopubs.org
e33
Information downloaded from jop.ascopubs.org and provided by at Charité - Med. Bibliothek on May 5, 2015 from 130.133.8.114 Copyright © 2015 American Society of Clinical Oncology. All rights reserved.
NOTE. Data in the first half of the table indicate patients in whom AC was ordered directly after medical oncology consultation without preceding postoperative complication and/or intercurrent illness; data for the full sample of 580 is shown in the second half of the table. Percentages may not add to 100% as a result of rounding. Abbreviations: AC, adjuvant chemotherapy; CVC, central venous catheter; IV, intravenous; MO, medical oncology; OR, odds ratio; Ref, reference. * Based on No. in row. † Patient comorbidity that would be considered relevant to fitness for chemotherapy. ‡ Unavailable for one patient.
Wasserman et al
those for which no compensation was received; those relationships marked with a “C” were compensated. For a detailed description of the disclosure categories, or for more information about ASCO’s conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.
Author Contributions Conception and design: David W. Wasserman, Christopher Booth, James J. Biagi
Corresponding author: James J. Biagi, Cancer Centre of Southeastern Ontario, Burr 2, Kingston General Hospital, 25 King St W, Kingston, ON, K7L 5P9 Canada; e-mail: [email protected].
DOI: 10.1200/JOP.2014.001531; published online ahead of print at jop.ascopubs.org on August 19, 2014.
References 1. Moertel CG, Fleming TR, Macdonald JS, et al: Levamisole and fluorouracil for adjuvant therapy of resected colon carcinoma. N Engl J Med 322:352-358, 1990 2. O’Connell MJ, Mailliard JA, Kahn MJ, et al: Controlled trial of fluorouracil and low-dose leucovorin given for 6 months as postoperative adjuvant therapy for colon cancer. J Clin Oncol 15:246-250, 1997 3. Andre´ T, Boni C, Navarro M, et al: Improved overall survival with oxaliplatin, fluorouracil, and leucovorin as adjuvant treatment in stage II or III colon cancer in the MOSAIC trial. J Clin Oncol 27:3109-3116, 2009 4. Yothers G, O’Connell MJ, Allegra CJ, et al: Oxaliplatin as adjuvant therapy for colon cancer: Updated results of NSABP C-07 trial, including survival and subset analyses. J Clin Oncol 29:3768-3774, 2011 5. Hershman D, Hall MJ, Wang X, et al: Timing of adjuvant chemotherapy initiation after surgery for stage III colon cancer. Cancer 107:2581-2588, 2006 6. Biagi JJ, Raphael MJ, Mackillop WJ, et al: Association between time to initiation of adjuvant chemotherapy and survival in colorectal cancer: A systematic review and meta-analysis. JAMA 305:2335-2342, 2011
e34
JOURNAL
OF
ONCOLOGY PRACTICE
•
7. Des Guetz G, Nicolas P, Perret GY, et al: Does delaying adjuvant chemotherapy after curative surgery for colorectal cancer impair survival? A meta-analysis. Eur J Cancer 46:1049-1055, 2010 8. Cancer Care Ontario: Cancer System Overview. https://www.cancercare.on. ca/ocs/csoverview/ 9. Central East Regional Cancer Program Statistics. www.ce-rcp.on.ca/ cancerstatistics/ 10. South East Regional Cancer Program Regional Cancer Plan 2011-2014. www.se-rcp.on.ca/common/pages/UserFile.aspx?fileId⫽117416 11. Cheung WY, Neville BA, Earle CC: Etiology of delays in the initiation of adjuvant chemotherapy and their impact on outcomes for stage II and III rectal cancer. Dis Colon Rectum 52:1054-1063, 2009 12. Lima IS, Yasui Y, Scarfe A, et al: Association between receipt and timing of adjuvant chemotherapy and survival for patients with stage III colon cancer in Alberta, Canada. Cancer [epub ahead of print on February 11, 2011] 13. Sargent D, Grothey A, Gray R: Time to initiation of adjuvant chemotherapy and survival in colorectal cancer. JAMA 306:1199, 2011
V O L . 11, I S S U E 1
Copyright © 2014 by American Society of Clinical Oncology
Information downloaded from jop.ascopubs.org and provided by at Charité - Med. Bibliothek on May 5, 2015 from 130.133.8.114 Copyright © 2015 American Society of Clinical Oncology. All rights reserved.
Employment or Leadership Position: None Consultant or Advisory Role: Rachel Goodwin, Sanofi (C), Novartis (C), Roche (C), Bayer (C) Stock Ownership: None Honoraria: None Research Funding: None Expert Testimony: None Patents, Royalties, and Licenses: None Other Remuneration: None
Administrative support: James J. Biagi Collection and assembly of data: David W. Wasserman, Majdi Boulos, Rachel Goodwin Data analysis and interpretation: David W. Wasserman, Wilma M. Hopman Manuscript writing: All authors Final approval of manuscript: All authors
Delays in Time to Adjuvant Chemotherapy
Appendix
CVC to AC (n = 98, 62, 245) MO Consult to CVC (n = 98, 62, 245) MO Referral to MO consultation Surgery to MO referral
10
8
6
4
2
0 Group I
Group II
Group III
Figure A1. Total TTAC and the time intervals among the three groups for all patient (n ⫽ 245) who had a CVC inserted. Group I includes patients who experienced a postoperative complication or intercurrent medical illness. Group II includes patients without a postoperative complication and/or intercurrent illness but in whom chemotherapy was deferred at time of medical oncology consultation due to either patient preference and/or need for further investigations. Group III includes patients in whom AC was ordered directly after MO consultation without preceding postoperative complication and/or intercurrent illness. AC, adjuvant chemotherapy; ECOG, Eastern Cooperative Oncology Group; MO, medical oncology; TTAC, time to adjuvant chemotherapy; VTE, venous thromboembolism.
Copyright © 2014 by American Society of Clinical Oncology
J A N U A R Y 2015
•
jop.ascopubs.org
e35
Information downloaded from jop.ascopubs.org and provided by at Charité - Med. Bibliothek on May 5, 2015 from 130.133.8.114 Copyright © 2015 American Society of Clinical Oncology. All rights reserved.
Time to Adjuvant Chemotherapy (weeks)
12
Original Contribution
Reasons for Delay in Time to Initiation of Adjuvant Chemotherapy for Colon Cancer By David W. Wasserman, MD, Majdi Boulos, MD, Wilma M. Hopman, MA, Christopher M. Booth, MD, Rachel Goodwin, MSc, MD, and James J. Biagi, MD
Abstract Purpose: Adjuvant chemotherapy (AC) improves survival among patients with colon cancer (CC). Two meta-analyses have demonstrated a decrease in survival with increasing time to AC (TTAC). Here, we examine the predominant factors leading to delay in TTAC.
Methods: Individual medical records of 580 patients with CC who initiated AC August 2005-November 2010 at two large academic cancer centers in Eastern Ontario were reviewed. Information regarding patient, disease, and treatment characteristics, including time intervals between each step in the cancer care pathway from surgery to AC, was captured. Patients were then categorized into three groups for comparison: (I) postoperative complication, (II) oncologist- or patient-initiated delay, (III) no delay. These groups were compared using 2 tests and one-way analysis of variance. A multivariable logistic regression model was used to determine factors associated with TTAC ⬎ 8 weeks in all patients and in group 1 alone.
Introduction In large phase III randomized clinical trials, adjuvant chemotherapy (AC) for colon cancer has been shown to significantly improve overall survival in stage III disease and trend toward increased disease-free survival in selected patients with stage II disease at high risk of relapse.1-4 In these studies, the first cycle of chemotherapy was administered within 8 weeks of surgery. However, in routine clinical practice, AC is often initiated more than 8 weeks after surgery.5 Two recent meta-analyses analyzing time to adjuvant chemotherapy (TTAC) in colon cancer have shown that the survival benefit of AC diminishes with increasing wait times after surgery.6-7 Our group has previously demonstrated that each 4-week increase in TTAC is associated with a significant decrease in both overall survival (hazard ratio [HR] ⫽ 1.14; 95% CI, 1.10 to 1.17) and disease-free survival (HR ⫽ 1.14; 95% CI, 1.10 to 1.18). Des Guetz et al7 used a dichotomous variable of 8 weeks and found that delaying AC for ⬎ 8 weeks was associated with worse overall survival (rate ratio ⫽ 1.20; 95% CI, 1.15 to 1.26). Few studies have examined the reasons underlying delays in TTAC. Here, we examine individual patients’ medical records from two large academic cancer centers in Ontario to determine reasons for delay that may not be captured with commonly used administrative data. We also analyze wait times along each step in the pathway of care from surgery to chemotherapy. We hypothesized that the majority of cases of TTAC ⬎ 8 weeks would e28
JOURNAL
OF
ONCOLOGY PRACTICE
•
Results: TTAC among the three groups was (I) 10.1 ⫾ 2.7 weeks, (II) 10.5 ⫾ 3.6 weeks, (III) 8.5 ⫾ 2.1 weeks (P ⬍ .001). The only significant predictor of TTAC ⬎ 8 weeks on multivariable analysis in group I was route of AC via central venous catheter (odds ratio [OR] ⫽ 2.4; 95% CI, 1.2 to 4.9). When multivariable analysis was performed on all patients, the presence of postoperative complications (OR ⫽ 2.4; 95% CI, 1.6 to 3.8) and oncologist- or patient-initiated delay were the strongest predictors of delay (OR ⫽ 3.5; 95% CI, 2.1 to 6.0). The percentages of patients with TTAC ⬎ 8 weeks were (I) 76.4% (n ⫽ 110), (II) 81.4% (n ⫽ 92), (III) 57.9% (n ⫽ 187). Conclusions: In patients with no reason for delay, most experienced TTAC ⬎ 8 weeks. This likely reflects delays in referral, consultation, and chemotherapy booking. These health-system factors are modifiable, and future quality improvement initiatives should focus on how to reduce them.
be due to health care system delays, specifically, the cumulative wait times between referrals, consultations, procedures, and finally chemotherapy. We also hypothesized that the minority of cases of TTAC ⬎ 8 weeks would be due to postoperative complications, physician-initiated delays, or patient-initiated delays.
Methods Study Population Ontario has the largest population of Canada’s provinces and territories. Universal health care is provided through a publically funded system, with a minor component of a parallel private system. Cancer care is overseen by Cancer Care Ontario and is administered through 14 local health integration networks (LHIN). Each LHIN has a regional comprehensive cancer center.8 There are six LHINs that contain academic medical centers. In these LHINs, the regional comprehensive cancer center is housed at the academic institution’s affiliated hospitals. We included two of the six academic regional cancer centers in Ontario for our study: The Ottawa Hospital Cancer Centre, with a catchment area of 1.6 million people,9 and the Cancer Centre of Southeastern Ontario in Kingston, Ontario, with a catchment area of 500,000 people.10 Appropriate research ethics board approval was obtained at both centers, and a datasharing agreement was completed. At both institutions, all chemotherapy is ordered through the software program OPIS
V O L . 11, I S S U E 1
Copyright © 2014 by American Society of Clinical Oncology
Information downloaded from jop.ascopubs.org and provided by at Charité - Med. Bibliothek on May 5, 2015 from 130.133.8.114 Copyright © 2015 American Society of Clinical Oncology. All rights reserved.
Queen’s University; Kingston General Hospital, Kingston; and University of Ottawa, Ottawa, Ontario, Canada
Delays in Time to Adjuvant Chemotherapy
2005. Using the OPIS 2005 chemotherapy administration database, patients with pathologically confirmed colon cancer who started AC between August 1, 2005 and November 30, 2010 were identified and had their medical records abstracted.
Data Abstraction
Copyright © 2014 by American Society of Clinical Oncology
Data were collected in a Microsoft Excel file and imported into IBM SPSS (version 20.0 for Windows) for statistical analyses. The time data were initially graphed to assess the underlying distribution. Data were then analyzed descriptively, including percentages and frequencies for categorical data (sex, comorbidities, surgical hospital, and stage); and means, standard deviations, medians, and quartiles for continuous data (time data, length of stay, and age). Analyses used the three aforementioned groups for comparison: Group I, Group II, and Group III. The groups were compared using 2 tests for categorical data, one-way analysis of variance for continuous data, and the Kruskal-Wallis test for length of stay. Factors associated with TTAC ⬎ 8 weeks were identified using 2 tests and independent samples t tests. Multivariable logistic regression was then used to assess which variables predict for TTAC ⬎ 8 weeks in all patients and in Group III group only. A two-sided P value ⬍ .05 was considered statistically significant.
Results Figure 1 shows the flow diagram of the 580 patients analyzed in our study. The number of patients categorized into the three groups was as follows: (I) 24.7% (n ⫽ 143), (II) 19.5% (n ⫽ 113), (III) 55.9% (n ⫽ 324). Therefore, three quarters (75.4%) of the patients had no medical or surgical reason for TTAC delay. The most common reason for delay in Group II was patient decision to delay treatment to consider whether he or she wanted to proceed with AC (n ⫽ 61, 54.0%). The most common postoperative complication in Group I was ileus (n ⫽ 49, 34.2%). Table 1 lists baseline patient characteristics and time data based on group categorization. As shown in Table 1 and Appendix Figure A1 (online only), total TTAC and the subintervals differed among the three groups. The only exception to this was the interval between surgeon referral and MO consultation, which was consistent among the three groups. Group III had the lowest TTAC, but the mean was still ⬎ 8 weeks (8.5 ⫾ 2.1 weeks). As would be expected, the time from MO consultation to AC was longest in Group II, and time from surgery to referral to MO was longest in Group I. Median length of stay in hospital after surgery in days was 10.0 (95% CI, 6.0 to 14.0), 6.0 (95% CI, 5.0 to 7.0), and 6.0 (95% CI, 4.0 to 7.0) for Group I, Group II, and Group III, respectively (P ⬍ .001, KruskalWallis). There were some notable differences in group characteristics. Oral chemotherapy was administered more frequently in Group II (37.2%), compared with Group I (23.8%) and Group III (17.9%; P ⫽ .001). There were more patients with comorbidity in Group I compared with Group II and Group III (50% v 39% and 38%, respectively; P ⫽ .054). There was no significant difference in mean age of the groups. Table 2 displays univariable and multivariable analysis of clinical, disease, and treatment factors associated with TTAC ⬎ 8 weeks using the 324 patients in Group III as well as all 580
J A N U A R Y 2015
•
jop.ascopubs.org
e29
Information downloaded from jop.ascopubs.org and provided by at Charité - Med. Bibliothek on May 5, 2015 from 130.133.8.114 Copyright © 2015 American Society of Clinical Oncology. All rights reserved.
All data abstraction was by performed by D.W.W. and M.B. in Kingston and Ottawa, respectively, coordinated with a test cohort and characterization by consensus to ensure consistency. The following data were collected from the patients’ electronic medical record (EMR): demographics, academic versus community surgical hospital, tumor stage, baseline prespecified comorbidities, length of stay in hospital after colon resection, postoperative complications or intercurrent medical illness, patients’ medical fitness for chemotherapy at time of medical oncology (MO) consultation, whether chemotherapy was ordered directly after initial MO consultation, why AC was not ordered directly after MO consultation (if applicable), route of chemotherapy administration, and time parameters. Time parameters collected included dates of surgery, referral to MO, MO consultation, central venous catheter (CVC) insertion (if applicable), and first cycle of chemotherapy administration. The MO consultation note served as the primary source of information for comorbidities and postoperative complications, with the discharge summary and EMR serving as secondary sources of information. Prespecified comorbidities for inclusion included cardiovascular, respiratory, renal, diabetes, and other. Patients were dichotomized into presence or absence of comorbidities, as well as presence or absence of postoperative complications. Patients’ fitness for chemotherapy at initial MO consultation was collected. Chemotherapy was determined to be ordered directly after initial consultation if the oncologist indicated that arrangements for chemotherapy would commence immediately. A patient was determined to be medically fit for chemotherapy unless explicitly stated otherwise in the consultation note. When chemotherapy was not ordered directly after the consult, the reason why it was not was determined from the consultation note. Subsequent clinic notes were examined if there was ambiguity about the reason from the consultation note. Patients were divided into three groups. Group I includes all patients who experienced a postoperative complication or intercurrent illness between surgery and MO consultation. Group II consists of patients with no postoperative complication or intercurrent medical illness, but in whom AC was deferred after MO consultation as a result of patient preference or the need for further investigations such as restaging. Group III includes patients without a postoperative complication or intercurrent illness between surgery and MO consultation who underwent AC directly after MO consultation without additional delay due to patient preference or need for further investigations. We considered splitting Group I into “fit for AC at time of MO consultation” and “not fit for AC at time of MO consultation” subgroups, but there would have been too few patients in the “not fit” subgroup for meaningful analysis.
Statistical Analyses
Wasserman et al
Initial search results (N = 640) Patients not eligible as a result of missing data (n = 60) Eligible patients (n = 580)
(Group II; n = 113) AC not ordered after consult Patient decision (n = 61) Restaging (n = 33) Evaluation for clinical trial (n = 13) Tumor board review (n = 8) Access to chemotherapy/approval (n = 6) Other (n = 15)
Postoperative complication or intercurrent medical illness (Group I) (n = 143)
(Group III; n = 324) AC ordered after consult
Medically fit for chemotherapy (n = 111) at time of MO consultation Ileus/SBO (n = 49) (n = 16) Wound infection Other infection (n = 14) Cardiovascular events (n = 10) Wound dehiscence (n = 10) VTE (n = 8) (n = 5) Pneumonia (n = 5) Anastomotic leak (n = 5) Postoperative bleeding Poor functional status (n = 5) leading to prolonged hospitalization/rehabilitation Other (n = 9)
Medically unfit for chemotherapy at time of MO consultation Wound healing/infection Nononcologic specialist follow-up Other infection ECOG > 2 Other
(n = 32) (n = 17) (n = 5) (n = 5) (n = 4) (n = 3)
Figure 1. Flow diagram of the patient population included in the study. AC, adjuvant chemotherapy; ECOG, Eastern Cooperative Oncology Group performance status; MO medical oncology; SBO, small bowel obstruction; VTE, venous thromboembolism.
patients. The only variable that was significantly associated with TTAC ⬎ 8 weeks in Group III was intravenous AC via CVC (odds ratio [OR] ⫽ 2.4; 95% CI, 1.2 to 4.9; P ⫽ .013). Notably, age and presence of comorbidities had no significant association with TTAC ⬎ 8 weeks. The strongest predictors of TTAC ⬎ 8 weeks in all 580 patients were the presence of a postoperative complication or intercurrent illness (OR ⫽ 2.4; 95% CI, 1.6 to 3.8; P ⱕ .001) and oncologist- or patientinitiated delay (OR ⫽ 3.5; 95% CI, 2.1 to 6.0; P ⱕ .001). There was no significant association for comorbidities, age, stage, or type of surgical hospital. Again, AC delivered IV with CVC was significantly associated with TTAC ⬎ 8 weeks compared with oral or IV without CVC (OR ⫽ 1.7; 95% CI, 1.0 to 2.9; P ⫽ .037). The percentages of patients with TTAC ⬎ 8 weeks in the three groups were (I) 76.9% (n ⫽ 110), (II) 81.4% (n ⫽ 92), and (III) 57.7% (n ⫽ 187). In other words, only 23.1% of Group I patients, 18.6% of Group II patients, and 42.3% of Group III patients received AC within 8 weeks. Importantly, there was no effect on TTAC if surgery was performed in an affiliated hospital versus a cancer center hospital (P ⫽ .32).
Discussion There are multiple available studies that have examined the association between delayed TTAC and survival in colorectal cancer. However, there is a paucity of data about the etiology of these delays. Two studies have examined administrative databases to answer this question.11,12 The first demonstrated that the presence of a postoperative complication or intercurrent e30
JOURNAL
OF
ONCOLOGY PRACTICE
•
illness was a significant predictor of increased TTAC with relation to rectal cancer adjuvant therapy in an American administrative database. They also observed that advanced age and black race were associated with greater delays to AC, but only in the group of patients without established contact with a medical oncologist before surgery. This paradigm of “established contact” is not applicable to colon cancer because medical oncologists do not see patients in consultation before surgery. We did not find any significant association between age and TTAC on multivariable analysis. Race was not examined as this information is not captured routinely in Canadian administrative health data sets. Others have previously suggested that comorbidity and postoperative course confound the relationship between delayed TTAC and decreased overall survival.13 They speculated that those who experience delayed TTAC are likely to have more comorbidity and are therefore more likely to have decreased survival. Our results demonstrate that TTAC is delayed among patients with postoperative complications. However the most notable findings of our study are (1) the majority of patients (75.4%) do not have any medical reason to delay AC, and (2) most of these patients (63.8%) have TTAC delayed ⬎ 8 weeks. The second study to investigate causes of delay in TTAC demonstrated, through administrative datasets in the province of Alberta, Canada, that patients who received chemotherapy ⬎ 12 weeks postsurgery had more comorbidities compared with patients who received AC within 12 weeks of surgery.12 Had their analysis been dichotomized at 8 weeks, it is uncertain whether there would have been a difference. They did not have
V O L . 11, I S S U E 1
Copyright © 2014 by American Society of Clinical Oncology
Information downloaded from jop.ascopubs.org and provided by at Charité - Med. Bibliothek on May 5, 2015 from 130.133.8.114 Copyright © 2015 American Society of Clinical Oncology. All rights reserved.
No postoperative complication or medical illness (n = 437)
Delays in Time to Adjuvant Chemotherapy
Table 1. Baseline Characteristics of 580 Patients Treated With Adjuvant Chemotherapy for Resected Colon Cancer at the Cancer Centre of Southeastern Ontario and the Ottawa Hospital, 2005 to 2010 Group I (n ⴝ 143) Characteristic
Mean
Age, years (range 23-90) No.
Group II (n ⴝ 113)
Group III (n ⴝ 324)
SD
Mean
SD
Mean
SD
10.7
65.2
11.9
62.9
10.8
No.
%
%
No.
%
Sex 85
59.4
60
53.1
168
51.9
58
40.6
53
46.9
156
48.1
Cancer Centre No
64
44.8
52
46.0
159
49.1
Yes
79
55.2
61
54.0
165
50.9
T classification T3
79
59.0
66
64.1
185
64.0
T4
55
41.0
37
35.9
103
35.6
N classification N0
35
24.5
30
26.5
54
16.7
N1
63
44.1
58
51.3
160
49.5
N2
45
31.5
25
22.1
109
33.7
Comorbidities No
72
50.3
69
61.1
201
62.0
Yes
71
49.7
44
38.9
123
38.0
Chemotherapy Oral IV with CVC IV without CVC
34
23.8
42
37.2
58
17.9
100
69.9
62
54.9
246
75.9
9
6.3
9
8.0
20
6.2
Chemotherapy ordered after consult No
57
39.9
113
100.0
0
0.0
Yes
86
60.1
0
0.0
324
100.0
Mean
SD
Mean
SD
Mean
SD
Time data, weeks Surgery to MO referral
3.6
⫾1.9
2.7
2.0
3.1
1.8
Referral to MO consultation
3.4
1.9
3.5
2.4
3.3
1.6
Consult to CVC*
2.5
1.7
4.0
3.6
1.9
1.1
1.4
1.0
1.4
0.6
No. CVC to chemotherapy*
98 0.8
No. Consult to chemotherapy Total time to adjuvant chemotherapy
62
98
245
62
0.7 245
3.2
2.4
4.7
3.4
2.2
1.3
10.1
2.7
10.5
3.6
8.5
2.1
NOTE. Group I comprises patients who experienced a postoperative complication or intercurrent medical illness; Group II comprises patients without a postoperative complication and/or intercurrent illness but in whom chemotherapy was deferred at time of medical oncology consultation as a result of either patient preference and/or need for further investigations; Group III comprises patients in whom adjuvant chemotherapy was ordered directly after MO consultation without preceding postoperative complication and/or intercurrent illness. Abbreviations: CVC, central venous catheter; IV, intravenous; MO, medical oncology. * CVC line insertion date not known for two patients in Group I and one patient in Group III.
access to data regarding postoperative complications, but they did have access to extensive demographic data. They reported that patients with TTAC ⬎ 12 weeks were more likely to live in neighborhoods with a high percentage of divorced, separated, or widowed; a low employment rate; and a low median household income. On average, in all three groups, the time between having surgery and MO consultation was between 6 and 7 weeks. This Copyright © 2014 by American Society of Clinical Oncology
leaves little time for a patient to decide on chemotherapy, complete restaging if necessary, obtain a CVC, and then book chair time in the chemotherapy suite. In patients without a postoperative complication or intercurrent illness, much greater effort should be made to have assessment by medical oncology sooner than is done in current practice. If time from surgery to MO consultation or time from referral to MO consultation could be decreased by an average of 1 week, then the average TTAC
J A N U A R Y 2015
•
jop.ascopubs.org
e31
Information downloaded from jop.ascopubs.org and provided by at Charité - Med. Bibliothek on May 5, 2015 from 130.133.8.114 Copyright © 2015 American Society of Clinical Oncology. All rights reserved.
Male Female
Wasserman et al
Table 2. Variables Associated With Initiation of Adjuvant Chemotherapy More Than 8 Weeks After Surgery for Patients With Colon Cancer Treated at the Cancer Centre of Southeastern Ontario and the Ottawa Hospital, 2005-2010 Proportion > 8 wk Characteristic
No.
Univariable Analysis
%*
OR
95% CI
Multivariable Analysis P
OR
95% CI
P
Patients in Whom AC Was Ordered Directly After MO Consult (n ⫽ 324) Patient related Age, years 93
55.7
Ref
Ref
65-74 (n ⫽ 108)
70
64.8
1.47
0.89 to 2.42
.13
1.62
0.95 to 2.75
.077
ⱖ 75 (n ⫽ 49)
25
51.0
0.83
0.44 to 1.57
.56
1.05
0.50 to 2.22
.89
Male (n ⫽ 168)
98
58.3
Female (n ⫽ 156)
90
57.7
0.97
0.63 to 1.52
.91
1.11
0.70 to 1.75
.67
1.15
0.73 to 1.82
.54
1.21
0.73 to 1.98
.46
Sex Ref
Ref
Relevant medical comorbidity† Yes (n ⫽ 123)
74
60.2
No (n ⫽ 201)
114
56.7
Ref
Ref
Disease related Stage‡ 2 (n ⫽ 54)
28
51.9
3 (n ⫽ 269)
159
59.1
105
63.6
83
Ref
Ref
0.75
0.41 to 1.34
.33
1.07
0.55 to 2.08
52.2
0.62
155
63.0
9
45.0
24
41.4
.83
0.40 to 0.97
.038
0.64
0.40 to 1.02
.061
2.41
1.35 to 4.32
.003
2.43
1.21 to 4.90
.013
1.16
0.42 to 3.23
.78
0.98
0.32 to 2.95
.97
Treatment related Surgical hospital Cancer Centre hospital (n ⫽ 165) Affiliate (n ⫽ 159)
Ref
Ref
Chemotherapy IV with CVC (n ⫽ 246) IV without CVC (20) Oral (n ⫽ 58)
Ref
Ref Results for All 580 Patients
Patient related Age, years ⬍ 65 (n ⫽ 285)
185
64.9
65-74 (n ⫽ 192)
133
69.3
1.22
0.82 to 1.80
.32
1.30
0.86 to 1.99
.22
ⱖ 75 (n ⫽ 103)
71
68.9
1.20
0.74 to 1.94
.46
1.35
0.76 to 2.38
.31
Male (n ⫽ 313)
209
66.8
Female (n ⫽ 267)
180
67.4
1.03
0.73 to 1.46
.87
1.14
0.79 to 1.64
.48
1.01
0.71 to 1.44
.95
0.97
0.66 to 1.43
.88
Ref
Ref
Sex Ref
Ref
Relevant medical comorbidity Yes (n ⫽ 238)
160
67.2
No (n ⫽ 342)
229
67.0
Ref
Ref
Disease related Stage‡ 2 (n ⫽ 119)
82
68.9
3 (n ⫽ 460)
306
66.5
Cancer Centre hospital (n ⫽ 305)
212
69.5
Affiliate (n ⫽ 275)
177
64.4
Ref 1.12
0.72 to 1.72
0.79
0.56 to 1.12
Ref .62
0.89
0.55 to 1.44
.19
0.83
0.58 to 1.20
.63
Treatment related Surgical hospital Ref
Ref .32
Reason for delay Postoperative complication (n ⫽ 143)
109
76.2
2.35
1.51 to 3.67
⬍.001
2.45
1.56 to 3.85
⬍.001
Oncologist/patient initiated (n ⫽ 113)
92
81.4
3.19
1.89 to 5.38
⬍.001
3.52
2.05 to 6.03
⬍.001
188
58.0
No reason (n ⫽ 324)
Ref
Ref continued on next page
e32
JOURNAL
OF
ONCOLOGY PRACTICE
•
V O L . 11, I S S U E 1
Copyright © 2014 by American Society of Clinical Oncology
Information downloaded from jop.ascopubs.org and provided by at Charité - Med. Bibliothek on May 5, 2015 from 130.133.8.114 Copyright © 2015 American Society of Clinical Oncology. All rights reserved.
⬍ 65 (n ⫽ 167)
Delays in Time to Adjuvant Chemotherapy
Table 2. (Continued) Proportion > 8 wk Characteristic
No.
%*
Univariable Analysis OR
95% CI
Multivariable Analysis P
OR
95% CI
P
Chemotherapy IV with CVC (n ⫽ 408)
280
68.6
1.22
0.81 to 1.84
.34
1.73
1.03 to 2.88
.037
IV without CVC (38)
23
60.5
0.86
0.41 to 1.79
.68
0.97
0.44 to 2.15
.94
Oral (n ⫽ 134)
86
64.2
Ref
Ref
would drop below an 8-week goal in patients with no justifiable reason for delay. Another important finding is that most patients who requested a delay to contemplate AC ended up deciding on oral chemotherapy rather than IV. We can see this relationship in Table 1, whereby the highest proportion of patients undergoing oral chemotherapy is in the Group II group. The highest proportion undergoing AC with CVC is in Group III. This may represent the more motivated patient subgroups that eagerly want to pursue the most efficacious treatment. On the basis of our review of individual patient medical records, we were able to examine time intervals along the continuum of care from time of surgery to start of chemotherapy, which previous studies were unable to do. Furthermore, we could identify Group II, which would not be ascertained from an administrative database. In our analysis, the two strongest factors that predict delayed TTAC are the presence of a postoperative complication or intercurrent illness and oncologistor patient-initiated delay, but as hypothesized, the absolute number of patients with TTAC ⬎ 8 weeks was highest in Group III, reflecting health system delays. The results of our study have led to practice changes locally, such as computerized physician notification of pathology results for patients with colorectal cancer, enhanced nurse navigator roles, and timely central line booking. There are limitations to our study. The first is that it was performed retrospectively. Second, we did not have access to socioeconomic data to examine whether socioeconomic factors played a role in our population. Third, there may be other contributing variables that we did not capture in our analysis, such total dose of chemotherapy administered or baseline serum carcinoembryonic antigen level. Fourth, our index of comorbidity was based on conditions reported in the surgical discharge and oncology clinics notes. Accordingly, it is possibly that comorbidity may be under-reported in this study. Fifth, because our study population is from a single, albeit large, geographical region, clinicians should be attentive when generalizing our results to their patient population. Lastly, we did not include patients seen by MO who declined chemotherapy altogether. However, there is no reason to suggest that the time to Copyright © 2014 by American Society of Clinical Oncology
seeing a medical oncologist in those patients would differ significantly from those who accept AC. In summary, our study indicates that the majority of patients with resected colon cancer do not have a clinical reason to justify delays in TTAC. The presence of comorbidities was uncommon in our population and when present had no relationship with TTAC ⬎ 8 weeks on multivariable analysis. The presence of a postoperative complication or intercurrent illness and oncologist- or patient-initiated delay are significantly associated with delays in TTAC ⬎ 8 weeks. Postoperative complication or intercurrent illness are beyond the control of a medical oncologist and are not easily modifiable. Oncologist- or patient-initiated delay is often unavoidable when a patient is unsure of whether or not to proceed with chemotherapy and requires time for careful consideration. Furthermore, we believe that if there is concern regarding metastases, restaging investigations should not be omitted solely to decrease a patient’s TTAC. Fortunately, these situations occur in the minority of patients with resected colon cancer. We should therefore direct our attention to the population of patients without any clinically identifiable reason for delay, which represent the majority of patients, but of whom only half initiate chemotherapy within 8 weeks from surgery. Examination of the time intervals along the pathway of care shows that there is room for improvement in referral from surgery to MO, time from referral to MO consultation, and from MO consultation to CVC insertion. Future quality improvement initiatives should identify strategies to decrease these health system delays. Our results do not allow us to define an optimum TTAC, but practically an 8-week maximum for most patients may be reasonable. Acknowledgment Presented in poster format at the 10th Gastrointestinal Cancers Symposium, San Francisco, CA, January 24-26, 2013. Authors’ Disclosures of Potential Conflicts of Interest Although all authors completed the disclosure declaration, the following author(s) and/or an author’s immediate family member(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a “U” are
J A N U A R Y 2015
•
jop.ascopubs.org
e33
Information downloaded from jop.ascopubs.org and provided by at Charité - Med. Bibliothek on May 5, 2015 from 130.133.8.114 Copyright © 2015 American Society of Clinical Oncology. All rights reserved.
NOTE. Data in the first half of the table indicate patients in whom AC was ordered directly after medical oncology consultation without preceding postoperative complication and/or intercurrent illness; data for the full sample of 580 is shown in the second half of the table. Percentages may not add to 100% as a result of rounding. Abbreviations: AC, adjuvant chemotherapy; CVC, central venous catheter; IV, intravenous; MO, medical oncology; OR, odds ratio; Ref, reference. * Based on No. in row. † Patient comorbidity that would be considered relevant to fitness for chemotherapy. ‡ Unavailable for one patient.
Wasserman et al
those for which no compensation was received; those relationships marked with a “C” were compensated. For a detailed description of the disclosure categories, or for more information about ASCO’s conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.
Author Contributions Conception and design: David W. Wasserman, Christopher Booth, James J. Biagi
Corresponding author: James J. Biagi, Cancer Centre of Southeastern Ontario, Burr 2, Kingston General Hospital, 25 King St W, Kingston, ON, K7L 5P9 Canada; e-mail: [email protected].
DOI: 10.1200/JOP.2014.001531; published online ahead of print at jop.ascopubs.org on August 19, 2014.
References 1. Moertel CG, Fleming TR, Macdonald JS, et al: Levamisole and fluorouracil for adjuvant therapy of resected colon carcinoma. N Engl J Med 322:352-358, 1990 2. O’Connell MJ, Mailliard JA, Kahn MJ, et al: Controlled trial of fluorouracil and low-dose leucovorin given for 6 months as postoperative adjuvant therapy for colon cancer. J Clin Oncol 15:246-250, 1997 3. Andre´ T, Boni C, Navarro M, et al: Improved overall survival with oxaliplatin, fluorouracil, and leucovorin as adjuvant treatment in stage II or III colon cancer in the MOSAIC trial. J Clin Oncol 27:3109-3116, 2009 4. Yothers G, O’Connell MJ, Allegra CJ, et al: Oxaliplatin as adjuvant therapy for colon cancer: Updated results of NSABP C-07 trial, including survival and subset analyses. J Clin Oncol 29:3768-3774, 2011 5. Hershman D, Hall MJ, Wang X, et al: Timing of adjuvant chemotherapy initiation after surgery for stage III colon cancer. Cancer 107:2581-2588, 2006 6. Biagi JJ, Raphael MJ, Mackillop WJ, et al: Association between time to initiation of adjuvant chemotherapy and survival in colorectal cancer: A systematic review and meta-analysis. JAMA 305:2335-2342, 2011
e34
JOURNAL
OF
ONCOLOGY PRACTICE
•
7. Des Guetz G, Nicolas P, Perret GY, et al: Does delaying adjuvant chemotherapy after curative surgery for colorectal cancer impair survival? A meta-analysis. Eur J Cancer 46:1049-1055, 2010 8. Cancer Care Ontario: Cancer System Overview. https://www.cancercare.on. ca/ocs/csoverview/ 9. Central East Regional Cancer Program Statistics. www.ce-rcp.on.ca/ cancerstatistics/ 10. South East Regional Cancer Program Regional Cancer Plan 2011-2014. www.se-rcp.on.ca/common/pages/UserFile.aspx?fileId⫽117416 11. Cheung WY, Neville BA, Earle CC: Etiology of delays in the initiation of adjuvant chemotherapy and their impact on outcomes for stage II and III rectal cancer. Dis Colon Rectum 52:1054-1063, 2009 12. Lima IS, Yasui Y, Scarfe A, et al: Association between receipt and timing of adjuvant chemotherapy and survival for patients with stage III colon cancer in Alberta, Canada. Cancer [epub ahead of print on February 11, 2011] 13. Sargent D, Grothey A, Gray R: Time to initiation of adjuvant chemotherapy and survival in colorectal cancer. JAMA 306:1199, 2011
V O L . 11, I S S U E 1
Copyright © 2014 by American Society of Clinical Oncology
Information downloaded from jop.ascopubs.org and provided by at Charité - Med. Bibliothek on May 5, 2015 from 130.133.8.114 Copyright © 2015 American Society of Clinical Oncology. All rights reserved.
Employment or Leadership Position: None Consultant or Advisory Role: Rachel Goodwin, Sanofi (C), Novartis (C), Roche (C), Bayer (C) Stock Ownership: None Honoraria: None Research Funding: None Expert Testimony: None Patents, Royalties, and Licenses: None Other Remuneration: None
Administrative support: James J. Biagi Collection and assembly of data: David W. Wasserman, Majdi Boulos, Rachel Goodwin Data analysis and interpretation: David W. Wasserman, Wilma M. Hopman Manuscript writing: All authors Final approval of manuscript: All authors
Delays in Time to Adjuvant Chemotherapy
Appendix
CVC to AC (n = 98, 62, 245) MO Consult to CVC (n = 98, 62, 245) MO Referral to MO consultation Surgery to MO referral
10
8
6
4
2
0 Group I
Group II
Group III
Figure A1. Total TTAC and the time intervals among the three groups for all patient (n ⫽ 245) who had a CVC inserted. Group I includes patients who experienced a postoperative complication or intercurrent medical illness. Group II includes patients without a postoperative complication and/or intercurrent illness but in whom chemotherapy was deferred at time of medical oncology consultation due to either patient preference and/or need for further investigations. Group III includes patients in whom AC was ordered directly after MO consultation without preceding postoperative complication and/or intercurrent illness. AC, adjuvant chemotherapy; ECOG, Eastern Cooperative Oncology Group; MO, medical oncology; TTAC, time to adjuvant chemotherapy; VTE, venous thromboembolism.
Copyright © 2014 by American Society of Clinical Oncology
J A N U A R Y 2015
•
jop.ascopubs.org
e35
Information downloaded from jop.ascopubs.org and provided by at Charité - Med. Bibliothek on May 5, 2015 from 130.133.8.114 Copyright © 2015 American Society of Clinical Oncology. All rights reserved.
Time to Adjuvant Chemotherapy (weeks)
12
