The Impact of Primary Care Physicians on the Surgical Presentation and Outcomes of Colorectal Cancer in Vulnerable Populations

Abstract

Background

Multiple socioeconomic and clinical factors have been implicated in the health disparities that exist amongst vulnerable populations with colorectal cancer. Efforts have been directed toward addressing these factors to improve outcomes. We evaluate the impact of primary care physicians (PCP) on the surgical presentation and outcomes of colorectal cancer at a safety-net hospital.

Methods

A retrospective chart review of 331 patients diagnosed with colorectal adenocarcinoma between 2014 and 2020 at a single-institution urban county medical center.

Results

The cohort was predominantly male (59%) and Hispanic (52.1%). Thirty-two percent of patients had a PCP at time of diagnosis. Patients with PCPs compared to those without PCPs had significantly lower rates of acute presentation (perforation or obstruction) (17.0 vs 38.1%, P < .001), higher rates of surgical resection (83.0 vs 70.7%, P = .016), and were less likely to have metastatic disease at presentation (20.4 vs 33.5%, P = .02). Overall, having a PCP also improved probability of survival (HR 1.36, P < .04).

Conclusion

Having a PCP at the time of colorectal cancer diagnosis is associated with improved outcomes in a safety-net population, with significant differences in surgical presentation and resection.

Keywords

primary care physician colorectal cancer outcomes emergency surgery safety-net population

Key Takeaways

• Patients with primary care physicians (PCPs) compared to those without PCPs had significantly lower rates of emergency presentation and higher rates of resection of primary tumor for colorectal cancer

• Patients with PCPs had improved probability of overall survival compared to those without PCPs

Introduction

Colorectal cancer remains one of the leading causes of cancer-related mortality in the United States, accounting for 8.7% of all cancer deaths in 2021.¹ Poorer outcomes for colorectal cancer disproportionately affect vulnerable populations, who face multiple barriers along the entire continuum of care, from diagnosis through treatment. Well known identified barriers include lack of insurance or access to health care, inadequate access to bowel preparation, and decreased health literacy, among others.^2,3 Patients who are treated at safety-net hospitals often present with advanced stages of colorectal cancer which limit or eliminate surgical treatment options, leading to higher rates of mortality.^4,5 The implementation of the Affordable Care Act (ACA) was designed to eliminate cost sharing for screening and to facilitate insurance coverage for these patients and has indeed had some success in improving outcomes.⁶ However, the effect on colorectal cancer screening and survival has been modest at best, and gaps still exist that need to be identified.⁷

One of these gaps may be the establishment of a primary care physician (PCP) for these patients. Despite increased insurance coverage, many patients continue to seek care only as needed or from the emergency department. However, PCPs are ideally situated to address many barriers to care, including facilitating screening and ensuring follow up to results. The objective of this study was to determine if the presence of a PCP at the time of colorectal cancer diagnosis was associated with improved outcomes in a safety-net hospital patient population and to examine the impact on surgical presentation of disease.

Methods

Study Design

Following exemption from the Institutional Review Board at Harbor-UCLA Medical Center, patients diagnosed with colorectal adenocarcinoma between 2014 and 2020 were identified by International Classification of Diseases (ICD)-10 codes (C18-C20). Adult patients with an isolated diagnosis of colorectal adenocarcinoma were included. Patients were excluded if an incorrect malignancy instead of colorectal adenocarcinoma was coded into the electronic medical record erroneously, if there was insufficient follow up, or if the diagnosis could not be confirmed by chart review.

Data were collected on patient demographics (sex, age, ethnicity), presence of a primary care physician at time of diagnosis, pathology reports (histopathologic findings, tumor biology), clinical stage at diagnosis and pathologic stage at time of surgery, as well as vital status.

Outcomes

The primary outcome was overall survival. Secondary outcomes included rates of emergency presentation, stage at diagnosis, and time to planned surgical and systemic therapies. Emergency presentation was defined as patients who first presented to the hospital with intestinal obstruction or perforation secondary to colorectal cancer requiring surgical intervention. In the statistical analysis of time to planned surgery, patients who presented needing emergency surgery were omitted. Cancer staging was based on the American Joint Committee on Cancer (AJCC) eighth edition classification system. Stage I and II malignancies were defined as local, Stage III as locally advanced, and Stage IV as metastatic.

Statistical Analysis

Continuous variables are shown as means with standard deviations and were analyzed using independent student’s t tests. Categorical variables are shown as percentages and were analyzed using Chi-squared tests. Bivariate analysis using Kaplan–Meier survival curve was used to assess long-term survival. Statistical significance was defined as P-value <.05. All statistical analysis was performed using SAS 9.4.

Results

384 patients were initially identified by ICD-10 code between 2014 and 2020. A total of 53 patients were subsequently excluded due to either incorrect coding or insufficient follow up within our system, yielding a total of 331 patients. The overall study population was predominantly male (59%, n=196), with a mean age 57.0 years (range: 22-89). Patients were also predominantly Hispanic (52.1%, n = 173), and 48.4% (n = 161) spoke Spanish as a primary language. Overall, 31.9% of patients had a PCP at the time of the colorectal cancer diagnosis (n = 106, Table 1).

Table 1.

Demographics Of Patients Diagnosed With Colorectal Cancer.

	PCP at Diagnosis (n = 106)	No PCP at Diagnosis (n=225)	P-value
Age (years)	59.2	56.0	.007
Sex, male	60 (56.6%)	136 (60.4%)	.507
Ethnicity	—		.413
Hispanic	54 (50.9%)	119 (52.9%)	—
White	19 (17.9%)	32 (14.2%)	—
Black	11 (10.4%)	23 (10.2%)	—
Asian	21 (19.8%)	40 (17.8%)	—
Other	1 (.9%)	11 (4.9%)	—
Diabetes mellitus	106 (31.1%)	225 (20.4%)	.033
Hypertension	48 (45.3%)	77 (34.2%)	.053
Hyperlipidemia	35 (33.0%)	27 (12.0%)	<.001
History of CAD/MI	15 (14.2%)	15 (6.7%)	.027
ESRD^a	1 (.9%)	0 (0%)	.320
Atrial fibrillation^a	3 (2.8%)	2 (.9%)	.332
History of tobacco use	38 (35.8%)	98 (43.6%)	.184
Obesity	47 (44.3%)	53 (23.6%)	<.001
Cirrhosis^a	2 (1.9%)	3 (1.3%)	.657
Anemia	51 (48.1%)	124 (55.1%)	.234

CAD/MI: coronary artery disease/myocardial infarction, ESRD (end stage renal disease).

^aFisher exact test.

Patients with a PCP at the time of cancer diagnosis had an improved probability of overall survival with a hazard ratio of 1.36, P = .04 on bivariate analysis (Figure 1). Overall, patients with a PCP had lower rates of emergency presentation, 17.0% vs 38.1% (OR 3.00, 95% CI: 1.7-5.3, P < .001). Intestinal obstruction was the more common cause of emergency presentation in both groups 15.1% vs 31.2%. Patients with a PCP also had higher rates of surgical resection (83.0 vs 70.7%, P = .016) and lower rates of metastatic disease at presentation, 20.4 vs 33.5% (OR 1.97, 95% CI: 1.1-3.4, P = .016) (Table 2). Patients with PCPs were more likely to have undergone screening colonoscopy (28.3% vs .5%, P < .001) as a method of diagnosis. The rates of diagnostic colonoscopy for either a positive fecal immunochemical test or symptoms were the same between groups.

Figure 1.

Overall survival for patients with a PCP at time of cancer diagnosis (red) vs patients without a PCP at time of cancer diagnosis (blue).

Table 2.

Clinical Staging And Presentation Characteristics.

	PCP at Diagnosis	No PCP at Diagnosis	Or (95% CI)	P-value
Stage at diagnosis	—			.050
Local	40 (38.8%)	76 (34.4%)	—
Locally advanced	42 (40.8%)	71 (32.1%)	—
Metastatic	21 (20.4%)	74 (33.5%)	—
Metastatic at presentation	21 (20.4%)	71 (33.5%)	1.97 (1.1-3.4)	.016
Presentation	—			<.001
Perforated	2 (1.9%)	15 (6.9%)	—
Obstructed	16 (15.1%)	68 (31.2%)
Screening colonoscopy	30 (28.3%)	1 (.5%)
Diagnostic colonoscopy	58 (54.7%)	134 (61.5%)
Acute presentation	18 (17.0%)	83 (38.1%)	3.00 (1.7-5.3)	<.001
Screening colonoscopy	30 (28.3%)	1 (.5%)	.12 (.002-.087)	<.001^a
Diagnostic colonoscopy	58 (54.7%)	134 (61.5%)	1.32 (.83-2.1)	.246

Fisher’s exact test.

At the time of cancer diagnosis, patients with a PCP had higher rates of diabetes mellitus, hyperlipidemia, and coronary artery disease when compared to patients without a PCP, 31.1 vs 20.4% (P = .033), 33.0 vs 12.0% (P < .001), and 14.2% vs 6.7% (P = .027), respectively. History of tobacco use was more common in patients without a PCP and obesity was more common in patients with a PCP. Rates of chronic obstructive pulmonary disease, end stage renal disease, atrial fibrillation, and cirrhosis were low and not significantly different between the two groups.

Patients in both groups presented with similar tumor biology with no significant differences seen in microsatellite instability (MSI), KRAS positivity, NRAS positivity, or BRAF positivity. There was also no difference between groups with regards to lymphovascular or perineural invasion on final pathological evaluation.

In terms of surgical treatment, patients with a PCP were more likely to have their colorectal cancer resected, 83.0 vs 70.7% (P = .016). With regards to medical treatments, there was no difference in either the proportions of patients who received adjuvant chemotherapy (PCP: 80.5% vs no PCP: 82.6%; P = .681), time to adjuvant chemotherapy (PCP: 68 days vs no PCP: 72 days; P = .546), or the number of cycles of adjuvant chemotherapy received (PCP: 7.11 vs no PCP: 6.84; P = .596). There was also no difference in times to medical or surgical treatments: surgical resection, neoadjuvant therapy, or palliative chemotherapy (Table 3).

Table 3.

Surgical And Medical Treatments For Patients With Or Without PCP At Time Of Cancer Diagnosis.

	PCP at Diagnosis	No PCP at Diagnosis	P-value
Primary resected	88/106 (83.0%)	159/225 (70.7%)	.016
Time to resection (d)	113.6	93.0	.192
Time to diversion (d)	26.5	45.2	.590
Time to neoadjuvant therapy (d)	79.3	72.4	.546
Time to adjuvant therapy (d)	67.7	72.8	.546
Time to palliative chemotherapy (d)	91.3	68.2	.332
Cycles of adjuvant chemotherapy	7.11	6.84	.596

Discussion

While other studies have postulated on the benefits of PCPs and cancer outcomes,⁸ the reasons for improved outcomes remain incompletely described. Possible contributing factors include improved screening, control of chronic medical conditions, and better coordination of multidisciplinary care.^9,10 In our present study, we find that the presence of a PCP leads to improved overall survival in patients with colorectal cancer, and that there are significant differences in the surgical presentation of disease.

The most striking contrast in the two study groups that may partially explain the difference in overall survival was the presentation of colorectal cancer. Patients with PCP were more than half as likely to present with a need for an emergent operation compared to those patients without a PCP. It is well established that emergent presentations are associated with poorer outcomes including higher rate of positive margin, poorer disease-free survival, and poorer overall survival in colorectal cancer patients.^11,12 It may be that PCPs obviate acute presentations by initiating or expediting earlier work up for patient symptoms. In our study, patients with PCPs are also more likely to undergo screening colonoscopy, which also likely impacts presentation.¹³

Surgical resection of the colorectal cancer primary was achieved more often in patients with a PCP, likely due to the earlier stage of presentation without presence of diffuse metastatic disease or a locally aggressive tumor. However, we did not find a difference in time to treatment, either surgery or chemotherapy between the groups, suggesting that the lower mortality with having a PCP was not necessarily related to improved coordination of care and more expeditious cancer treatment. It is true that our particular health care system, historically having a predominantly uninsured population, has developed pathways to expedite and facilitate care in patients who receive a pathologic diagnosis of colorectal cancer. While the impact of having a PCP on time to cancer treatment is an understudied area, at least several studies have shown a modest decrease in time interval to diagnostics or treatment for patients with PCP.^14,15

Ironically, although patients with a PCP at diagnosis were found to have improved overall survival, they also had higher rates of concomitant comorbidities, including diabetes mellitus, hyperlipidemia, and coronary artery disease. We suspect that patients without PCPs were underdiagnosed with these conditions. Furthermore, the management of chronic diseases is also clearly an important aspect that may influence mortality. Uncontrolled diabetes and hypertension are well known to be associated with increased mortality and cardiovascular disease.^16,17 Obesity and history of tobacco use were significantly more prevalent in the patients without PCP, likely related to the absence of preventive care.

In our study, we were unable to evaluate insurance status at the time of presentation, which is a significant limitation. Higher quality insurance has been shown to have a positive correlation with access to PCPs and poorer quality insurance or lack of insurance has been associated with poorer colorectal cancer outcomes regardless of race.^18,19 It is interesting to note that even in patients who had PCPs at diagnosis, a fair number still presented with advanced disease, with 20% presenting with metastatic disease and 15% with obstruction. Screening colonoscopies, although higher in patients with PCPs, were still overall very low. As such, while we demonstrate that having a PCP is beneficial in many ways, the adherence to and continuity of care likely remain areas for improvement.

Conclusion

The presence of a PCP at the time of colorectal cancer diagnosis was associated with improved overall survival, decreased proportions of patients with emergency presentations and metastatic disease and higher proportions of patients undergoing surgical resection and screening colonoscopy.

Footnotes

Author Contributions

Naveen Balan: study design, data collection, data analysis, article writing and editing

Caitlyn Braschi: data collection, data analysis, article editing

Patrick Kirkland: study design, article editing

Amy H. Kaji: data analysis, article editing

Kathryn T. Chen: study design, data analysis, article writing and editing

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Naveen Balan

Caitlyn Braschi

References

Cancer Stat Facts: Colorectal Cancer; 2022. (Accessed March 20, 2022).https://seer.cancer.gov/statfacts/html/colorect.html

McDougall

Blair

Wiggins

, et al. Socioeconomic disparities in health-related quality of life among colorectal cancer survivors. J Cancer Surviv. 2019;13(3):459-467. doi:10.1007/S11764-019-00767-9

Olaku

Taylor

. Cancer in the medically underserved population. Prim Care. 2017;44(1):87-97. doi:10.1016/J.POP.2016.09.020

Lam

ATT

Kouzminova

Lin

. Characteristics of colorectal cancer survival in an urban county hospital. J Gastrointest Cancer. 2013;44(1):68-72. doi:10.1007/S12029-012-9446-6

Kornfield

Vittinghoff

Inadomi

Yee

Somsouk

. Late presentation of colorectal cancer in a vulnerable population. Am J Gastroenterol. 2013;108(4):466-470. doi:10.1038/AJG.2012.256

Kim

Koroukian

Statler

Rose

. The effect of Medicaid expansion among adults from low-income communities on stage at diagnosis in those with screening-amenable cancers. Cancer. 2020;126(18):4209-4219. doi:10.1002/CNCR.32895

Kelly

AMB

Kushi

Reed

Koh

Spiegelman

. Impact of the affordable care act on colorectal cancer outcomes: A systematic review. Am J Prev Med. 2020;58(4):596-603. doi:10.1016/j.amepre.2019.11.018

Ferrante

. Primary care utilization and colorectal cancer outcomes among medicare beneficiaries. Arch Intern Med. 2011;171(19):1747. doi:10.1001/archinternmed.2011.470

Triantafillidis

Vagianos

Gikas

Korontzi

Papalois

. Screening for colorectal cancer: the role of the primary care physician. Eur J Gastroenterol Hepatol. 2017;29(1):e1-e7. doi:10.1097/MEG.0000000000000759

10.

Lawrence

McLoone

Wakefield

Cohn

. Primary care physicians’ perspectives of their role in cancer care: A systematic review. J Gen Intern Med. 2016;31(10):1222-1236. doi:10.1007/S11606-016-3746-7

11.

Dekker

Tanis

Vleugels

JLA

Kasi

Wallace

. Colorectal cancer. Lancet. 2019;394(10207):1467-1480. doi:10.1016/S0140-6736(19)32319-0

12.

Belt

EJT

Stockmann

HBAC

Abis

GSA

, et al. Peri-operative bowel perforation in early stage colon cancer is associated with an adverse oncological outcome. J Gastrointest Surg. 2012;16(12):2260-2266. doi:10.1007/S11605-012-2053-9

13.

Boguradzka

Wiszniewski

Kaminski

, et al. The effect of primary care physician counseling on participation rate and use of sedation in colonoscopy-based colorectal cancer screening program--a randomized controlled study. Scand J Gastroenterol. 2014;49(7):878-884. doi:10.3109/00365521.2014.913191

14.

Provost

Pineault

Tousignant

Roberge

Tremblay

Breton

, et al.. Does the primary care experience influence the cancer diagnostic process? Int J Family Med. 2015;2015:12. doi:10.1155/2015/176812

15.

Walsh

Lofters

Moineddin

Krzyzanowska

Grunfeld

. Primary care continuity and wait times to receiving breast cancer chemotherapy: A population-based retrospective cohort study using CanIMPACT Data. Curr Oncol. 2021;28(6):4786-4804. doi:10.3390/curroncol28060405

16.

Navarro-Pérez

Orozco-Beltran

Gil-Guillen

, et al. Mortality and cardiovascular disease burden of uncontrolled diabetes in a registry-based cohort: the ESCARVAL-risk study. BMC Cardiovasc Disord. 2018;18(1):180. doi:10.1186/S12872-018-0914-1

17.

Cushman

. The burden of uncontrolled hypertension: Morbidity and mortality associated with disease progression. J Clin Hypertens. 2003;5(3):14-22. doi:10.1111/J.1524-6175.2003.02464.X

18.

Pulte

Jansen

Brenner

. Social disparities in survival after diagnosis with colorectal cancer: Contribution of race and insurance status. Cancer Epidemiol. 2017;48:41-47. doi:10.1016/J.CANEP.2017.03.004

19.

Alcalá

Roby

Grande

McKenna

Ortega

. Insurance type and access to health care providers and appointments under the affordable care act. Med Care. 2018;56(2):186-192. doi:10.1097/MLR.0000000000000855