Potential for Reducing Time to Referral for Colorectal Cancer Patients in Primary Care

Nicole F. van Erp; Charles W. Helsper; Saskia M. Olyhoek; Ramon R. T. Janssen; Amber Winsveen; Petra H. M. Peeters; Niek J. de Wit

doi:10.1370/afm.2446

Article Figures & Data

Figures

Tables
Additional Files

Download figure
Open in new tab
Download powerpoint
Figure 1
Flowchart for CRC patient selection.
CRC = colorectal cancer; GP = general practitioner.
Download figure
Open in new tab
Download powerpoint
Figure 2
Factors contributing to longest duration (≥219 days) for symptomatic CRC patients.^a
CRC = colorectal cancer; RBL = rectal blood loss.
^aThe Venn diagram circles do not illustrate the relative sizes of groups.

View popup

Table 1

Disease Characteristics of Symptomatic CRC Patients Referred by GP (N = 309)

Characteristic	No. (%)
TNM stage at diagnosis
0	1 (0.3)
I	41 (13.3)
II	83 (26.9)
III	114 (36.9)
IV	65 (21.0)
Unknown	5 (1.6)
Tumor location
Proximal colon	90 (29.1)
Distal colon	20 (6.5)
(Recto)sigmoid	99 (32.0)
Rectal	97 (31.4)
Colon unspecified	3 (1.0)

CRC = colorectal cancer; GP = general practitioner; TNM = tumor, nodes, metastases.

View popup

Table 2

Patient and Presentation Characteristics of Symptomatic CRC Patients, Duration of Primary Care Interval, and Log-Binomial Regression Analysis for 75th Percentile (N = 309)

Characteristic	No.	Duration, d, Median (IQR)	P Value^a	P90^b	Univariable RR (95% CI)	Multivariable RR (95% CI)
Age, y
≤50	35	34 (1-233)		491	1.5 (0.8-3.0)
51-60	47	3 (1-15)		408	0.6 (0.3-1.4)
61-70	100	14 (1-47)		94	0.8 (0.4-1.4)
71-80	91	6 (1-61)		204	0.9 (0.5-1.7)
81-90	36	8 (1-68)	.154	150	1 (ref)
Sex
Male	155	4 (1-47)		101	1 (ref)	1 (ref)
Female	154	13 (1-78)	.004	321	1.6 (1.1-2.4)	1.4 (0.9-2.1)
SES 2010^c
Low	81	12 (1-72)		240	1 (ref)
Medium-low	79	9 (1-63)		239	0.9 (0.5-1.5)
Medium-high	73	7 (1-51)		118	0.7 (0.4-1.3)
High	76	6 (1-47)	.551	223	0.7 (0.4-1.2)
Registered comorbidity^d
Chronic somatic
No	62	5 (1-48)		326	1 (ref)
Yes	247	10 (1-61)	.317	198	1.1 (0.7-1.9)
≥2	181	9 (1-58)		203	1.1 (0.6-1.8)
≥4	70	8 (1-43)		98	0.9 (0.5-1.7)
Gastrointestinal
No	256	8 (1-63)		219	1 (ref)
Yes	53	15 (2-48)	.622	119	0.7 (0.4-1.3)
Psychiatric
No	290	8 (1-58)		204	1 (ref)
Yes	19	22 (2-84)	.203	538	1.1 (0.5-2.3)
Registered family history of CRC^e
Not registered	267	11 (1-65)		233	1 (ref)
Negative	30	1 (1-13)		87	0.5 (0.2-1.3)
Positive	12	2 (1-34)	.003	87	0.3 (0.1-2.0)
Consultation frequency for year prior to first CRC consultation
≤2	56	2 (1-29)		117	1 (ref)
3-11	188	11 (1-64)		235	1.4 (0.8-2.5)
≥12	65	12 (1-54)	.093	120	1.2 (0.6-2.3)
History of malignancy
No	267	7 (1-50)		219	1 (ref)	1 (ref)
Yes	42	18 (2-84)	.101	178	1.7 (1.1-2.6)	1.5 (0.9-2.2)
Main registered symptom at first CRC consultation^f
Alarm GI symptom(s)	168	2 (1-28)		123	1 (ref)	1 (ref)
Nonalarm GI symptom(s)	113	26 (5-87)		257	1.9 (1.2-2.8)	1.7 (1.1-2.6)
Other symptom(s)	28	13 (2-43)	.000	273	0.9 (0.4-2.2)	0.9 (0.4-2.1)
Hemorrhoids at physical examination^g
No	298	8 (1-54)		219	1 (ref)
Yes	11	69 (1-115)	.192	213	2.3 (1.3-4.1)
TNM stage at diagnosis
0	1	87		…	…
I	41	2 (1-42)		83	1 (ref)
II	83	7 (1-48)		213	1.1 (0.5-2.3)
III	114	7 (1-48)		159	1.1 (0.5-2.2)
IV	65	23 (3-92)		502	1.9 (0.9-3.8)
Unknown	5	5 (1-246)	.013	…	…

CRC = colorectal cancer; GI = gastrointestinal; GP = general practitioner; IPC = primary care interval; IQR = interquartile range; P75 = 75th percentile value of the duration distribution; P90 = 90th percentile value of the duration distribution; RR = relative risk; SES = socioeconomic status score; TNM = tumor, nodes, metastases.
↵a P values based on Mann-Whitney U tests for variables with 2 categories and Kruskall-Wallis tests for variables with 3 or more categories.
↵b P90 value = 90th percentile value of the duration distribution; that is the IPC duration time in days where 90% of the population was below and 10% above.
↵c Socioeconomic status scores 2010 were retrieved from publicly available data from the Netherlands Institute for Social Research.16 Lowest SES score was defined as: SES score of <1 SD than the Dutch mean of 2010, Medium-low: 1 SD to mean SES score, Medium-high: mean SES score to +1 SD and Highest: > +1 SD higher than Dutch mean.
↵d Chronic somatic comorbidities were defined according to O’Halloran et al.17 Gastrointestinal comorbidities were all relevant GI-related registered comorbidities or conditions in either episode list or mentioned during GP consultations: irritable bowel syndrome, reflux disease, esophagitis, dyspepsia, abdominal pain, peptic ulcer, hiatus or abdominal hernia, benign GI neoplasms/polyps, constipation, chronic diarrhea, cholelithiasis, diverticulosis, anal fissures. Psychiatric comorbidities were all chronic psychiatric comorbidities according to O’Halloran et al.17
↵e Registered occurrence of colorectal cancer in a first degree family member.
↵F Alarm symptoms for colorectal cancer were defined as rectal blood loss, unintended weight loss, anemia, and a palpable tumor. GI symptoms include all GI-related, nonalarming symptoms. Other symptoms are all remaining, nonalarming, non-GI symptoms.
↵g Multivariable model excludes this factor due to low patient numbers.

Additional Files

Figures
Tables

Supplemental Appendix

Supplemental appendix

Files in this Data Supplement:

Supplemental data: Appendix - PDF file
The Article in Brief

Potential for Reducing Time to Referral for Colorectal Cancer Patients in Primary Care

Charles W. Helsper , and colleagues

Background Delay in referral to specialty care for patients who have symptoms of colorectal cancer may lead to poor health outcomes. While the current median duration from first consultation with cancer related complaints to referrals is approximately one week for colorectal cancer patients in the UK and the Netherlands, time to referral varies strongly. Some cases document months and even years before referral, and explanations for suboptimal referral is often incomplete or oversimplified.

What This Study Found A retrospective cohort study in the Netherlands reviewed the time to specialty referral for a group of 309 patients with colorectal cancer who initially presented with symptoms to their primary care doctor. A multivariable analysis narrowed down the factors associated with delays in referrals to specialty care. Those who initially presented with red flag symptoms, such as rectal bleeding or unintended weight loss, experienced shorter wait time than those who presented with non-alarming gastrointestinal symptoms. Univariable analysis showed that female patients and patients without a registered family history of the disease were also more likely to have a longer wait period. Of the 10% of patients with the longest wait times for referral to specialty care all patients had received an alternative initial diagnosis from their primary care physicians. These patients usually presented with conditions that obscured concern for colorectal cancer, such as hemorrhoids, fissures and inflammatory bowel disease. Initial diagnoses were not always reconsidered when complaints persisted and follow-up consultations were sometimes omitted.

Implications

This study points to a relationship between long time to referral of colorectal cancer in primary care and low cancer suspicion. There is potential for reducing longest times to referral of colorectal cancer patients in primary care by earlier reconsideration of the initial hypothesis and implementing strict follow-up consultations.
Visual Abstract

Files in this Data Supplement:

Adobe PDF - visual_abstract_van_erp_sept_2019.png.pdf