Complications Associated With Use of Long-Term Central Venous Catheters Among Commercially Insured Women With Breast Cancer

Long-term central venous catheter (CVC) use is associated with increased morbidity. Evidence-based guidelines tools may help decrease discretionary use of long-term CVCs resulting from provider- or institution-driven variations in practice.

Abstract

Purpose:

Despite some advantages to their use, long-term central venous catheters (CVCs) are associated with complications for patients who require chemotherapy. Understanding of these risks in commercially insured populations is limited. This information can inform medical policies that ensure the appropriate use of venous access devices. This study's objectives were to assess the extent of variation in use of long-term CVCs in a cohort of commercially insured women with breast cancer, and to assess risks of associated complications.

Methods:

Retrospective cohort analysis was conducted using health insurance claims between January 2006 and October 2013. The cohort included commercially insured women age ≥ 18 years diagnosed with breast cancer who received infusion chemotherapy (N = 31,047). We conducted matched and case-mix adjusted Cox proportional hazard modeling to assess differences in bloodstream infections and thrombovascular complications between patients using long-term CVCs and those using temporary intravenous catheters.

Results:

Approximately two thirds of the cohort had a long-term CVC, although rates varied across regions (57% to 75%), health plans (65% to 70%), and insurance coverage (63% to 68%). After propensity score matching, the adjusted hazard ratio for infection was 2.70 (95% CI, 2.31 to 3.16) and thrombovascular complications, 2.61 (95% CI, 2.33 to 2.93) in patients with long-term CVCs compared with those with temporary intravenous catheters.

Conclusion:

Although long-term CVCs may have benefits, they are associated with increased morbidity. Regional and health plan variation in long-term CVC insertion suggests that some of their use reflects provider- or institution-driven variation in practice. Evidence-based guidelines and tools may help decrease discretionary use of long-term CVCs.

Introduction

Long-term central venous catheters (CVCs) offer benefits to patients who receive chemotherapy. However, the use of CVCs may result in mechanical, infectious, and thrombotic complications,^1–5 especially in patients with cancer who may have additional risk factors for complications.^5–8

Patients with breast cancer often require venous access for administration of chemotherapy, antibiotics, and other therapies. Intravenous (IV) access is commonly obtained by using an IV catheter that is temporarily inserted into a vein and removed after treatment. An alternative is to use a long-term CVC, which can facilitate the delivery of infusion therapy and decrease patients' discomfort and anxiety associated with repeated IV insertions.^1,3,9 Temporary IV catheters carry an extravasation risk and can also take nurses added time to insert for each administration.¹⁰ Long-term CVCs offer patients and clinicians ready, convenient venous access and can support continuous chemotherapy infusion. They are largely accepted by patients who have them, especially those who do not experience complications.^11–13 The most commonly used long-term CVCs include surgically implanted cuffed tunneled CVCs, subcutaneous implanted ports (eg, port-a-cath), peripherally inserted CVCs, and percutaneous noncuffed or tunneled catheters.⁹

Despite the benefits, especially for administration of vesicant chemotherapy medications, there are well-recognized risks of complications associated with the use of long-term CVCs. The magnitude of this risk is not well defined, and there are few data regarding the risks associated with the different types of venous access devices for use during administration of chemotherapy.^1,14 Guidance regarding the appropriate mechanism for chemotherapy administration is lacking. The recent ASCO guidelines on CVC care for patients with cancer highlighted the need for additional research in this area.¹

Population-based assessments can build on existing knowledge by identifying patterns of use and outcomes across different types of patients and settings of care. A recent study using population-based SEER-Medicare data found a substantially increased risk of bloodstream infections associated with the use of long-term CVCs among patients over age 65 years with one of six types of cancer; the risk among patients with breast cancer was six-fold greater.¹⁵ Because the median age of diagnosis at breast cancer is 61, and older women are less likely to receive chemotherapy, the results of this study may not be generalizable to the majority of women with breast cancer who may be considered for a CVC.¹⁶

Given the high observed infection risk in older adults with breast cancer and health plans' business priority to examine care management for members with breast cancer, we conducted this study to focus on complications of long-term CVCs in a commercially insured population with breast cancer. This group included women mostly under the age of 65 years. The objectives of this study were two-fold: (1) to identify factors associated with variation in the use of long-term CVCs in women with breast cancer receiving chemotherapy, and (2) to assess the risk of infectious and thrombovascular complications associated with the use of long-term CVCs compared with temporary IV catheters in this population.

Methods

We conducted a retrospective cohort analysis using health insurance claims between January 2006 and October 2013 for a commercially insured population with breast cancer treated with injectable chemotherapy drugs. We compared complications between patients with long-term CVCs and patients with temporary IV catheters.

Data Source

The data for this study were obtained from the HealthCore Integrated Research Database, which contains a broad, clinically rich and geographically diverse spectrum of longitudinal medical and pharmacy claims data from 14 large commercial health plans in the United States. This database represents nearly 60 million members with medical or pharmacy coverage, including Medicare Advantage and supplemental insurance plans, and includes claims information from one of the largest commercially insured populations in the United States.

Both medical and pharmacy claims and member eligibility information from January 1, 2006, through October 31, 2013, were used for this study. All claims data were from a limited data set with de-identified patient information; strict measures were taken to preserve anonymity and confidentiality, and to ensure full compliance with the Health Insurance Portability and Accountability Act. No patients were directly involved in the study; therefore, this study was exempt from institutional review board review.

Study Cohort

The study cohort consisted of women age 18 years and over diagnosed with breast cancer who received an injectable form of chemotherapy to treat their disease. To be eligible, members needed at least two claims for medical care with a diagnosis of breast cancer (International Classification of Diseases, ninth edition [ICD-9] diagnosis codes: 174.0 to 174.9), or one claim with a breast cancer diagnosis and a claim for a relevant surgical procedure (mastectomy, lumpectomy, lymph node surgery) or treatment (external- beam radiation ootherapy r brachytherapy) between January 1, 2007, and October 31, 2012. We also required at least one claim for an injectable chemotherapy drug on or after the date of the first breast cancer diagnosis claim. (Appendix Table A1, online only) Exclusion criteria included: less than 12 months of continuous medical eligibility before the index date, a diagnosis for end-stage renal disease or renal failure, and a procedure indicating the presence of hemodialysis during the study period.

The variable of interest was use of a long-term CVC, defined by a medical claim for insertion of a long-term CVC. Patients with no claim for CVC insertion were assumed to have received chemotherapy through a temporary IV catheter. Long-term CVCs included implanted catheters (eg, ports or pumps), tunneled catheters (eg, Hickman catheters), and peripherally inserted central catheters (Appendix Table A1) For each user of a long-term CVC, the study index date was defined as the date of the first CVC claim after the first claim with a breast cancer diagnosis. For each patient without a claim for a long-term CVC insertion, the study index date was the date of the first claim for injectable chemotherapy.

Outcomes

The primary outcome was complications potentially related to catheter use. On the basis of a literature review and clinician input, we defined two complications, or “venous complications” for this study: bloodstream infections, and thrombovascular complications.^2,3,15,17 The end point was limited to complications occurring during a hospital stay or emergency department visit. Complications documented in the office or outpatient setting were not included in order to limit the analysis to the most severe complications and to minimize possible double-counting of events that were follow-up visits.

Bloodstream infections were identified by ICD-9 diagnosis codes for septicemia, bacteremia, sepsis, infection, and inflammatory reaction to vascular device, and other infection after infusion/injection. Thrombovascular complications included major bleeding, thrombosis, embolism, and thromboembolism (Appendix Table A1) We counted the first occurrence of each complication after the study index date as the primary outcome. In some cases, it was difficult to determine whether a complication was directly related to the catheter, given the lack of granularity associated with the ICD-9 codes. Therefore, we included a comparison group of patients without long-term CVCs in our analysis to account for complications unrelated to CVC use, which we would assume to be similar between groups.

Covariates

Covariates included patient demographic characteristics, comorbidities, prior venous complications, disease characteristics, cancer treatment, and total health care costs. Demographic characteristics included age, region (according to US Census regions and divisions),¹⁸ rural residence, health plan, line of business, enrollment in a Medicare plan (Medicare Advantage or supplement). Members were defined as living in a rural area if their residence was located outside of a metropolitan statistical area (MSA).¹⁹ We used a modified Deyo-Charlson Index, excluding metastatic solid tumor, to define the presence of comorbidities during the 12-month preindex period.²⁰ We also identified vascular complications (ie, same events that could be complications of a venous catheter) during this preindex period. Other disease characteristics included the presence of neutropenia, anemia, and metastases. Surgery, radiation therapy, and brachytherapy were identified for both the pre- and postindex periods to describe and account for cancer-directed treatments. In addition, because trastuzumab is typically prescribed for 1 year after surgery, we specifically identified use of this therapy. Lastly, we estimated members' total health care costs from the preindex period by summing the paid amounts on their medical claims. We adjusted the total costs to 2013 dollars using the medical care consumer price index (CPI).²¹

Analysis

Using bivariate analyses, we assessed factors associated with the use of CVCs among the entire cohort (N = 31,047). We used logistic regression to estimate the odds of having a long-term CVC, adjusting for patient demographic, disease, and treatment characteristics. In order to examine venous complications associated with use of a long-term CVC, we matched each patient with a long-term CVC to a patient with a temporary IV catheter using propensity score matching.²² We accounted for age, year of index date, health plan region and type, metastatic cancer, Medicare status, Deyo-Charlson Index, breast cancer surgery and radiation therapy or brachytherapy, as well as venous complications and health care costs from the preindex period. The final sample for the primary analysis included 10,483 matched pairs.

We used Cox proportional hazards regression to assess the impact of use of a long-term CVC on risk of complications. We modeled the outcomes separately: bloodstream infection, and thrombovascular complications. In each model, use of a long-term CVC was included as a binary variable. We adjusted for demographic and health insurance characteristics, comorbidities, prior venous complications, disease characteristics, breast cancer treatments, total health care costs in the previous year, and index year. We performed several sensitivity analyses, including different combinations of variables in the regression models such as health plan type, type of insurance, coordination of benefits, and median household income in zip code of residence. Based on a priori hypotheses, we also evaluated potential interaction effects of the use of a long-term CVC with age, metastatic disease, or a preindex venous complication.

Analyses were conducted using SAS software (version 9.2; SAS Institute, Cary, NC). A two-sided P value less than .05 was considered to be statistically significant.

Results

The sample consisted of 31,047 women with breast cancer who received chemotherapy. Approximately two thirds of the women had a claim for insertion of a long-term CVC during the study period. Patients with long-term CVCs were slightly younger that those with temporary IV catheters (mean, 55.4 years [standard deviation, 11.4 years] v 58.3 years [standard deviation, 12.9 years]; P < .001) and were more likely to have received treatment with trastuzumab at some point during their history (Table 1). There was no difference in comorbidities between patients with a long-term CVC and those without. After propensity score matching, there were no significant differences in the characteristics of those who received a long-term CVC compared with those who received a temporary IV catheter.

Table 1.

Cohort Characteristics, by Long-Term CVC Use

Characteristic	Before Match*		After Match†
	Long-Term CVC, No. (%)	Temporary IV Catheter, No. (%)	Long-Term CVC, No. (%)	Temporary IV Catheter, No. (%)
No. of patients	20,377	10,670	10,483	10,483
Age in years, mean (SD)	55.4 (11.4)	58.3 (12.9)	58.0 (11.3)	58.1 (12.8)
18-50	7,114 (34.9)	3,121 (29.3)	2,768 (26.4)	3,120 (29.8)
51-64	9,377 (46.0)	4,623 (43.3)	5,045 (48.1)	4,588 (43.8)
65-80	3,458 (17.0)	2,205 (20.7)	2,353 (22.5)	2,106 (20.1)
≥ 81	428 (2.1)	720 (6.8)	317 (3.0)	699 (6.4)
Residence region
Northeast	3,829 (18.8)	2,897 (27.2)	2,497 (23.8)	2,834 (27.0)
Midwest	5,459 (26.8)	2,149 (20.1)	2,311 (22.1)	2,143 (20.4)
South	6,641 (32.6)	2,275 (21.3)	2,678 (25.6)	2,266 (21.6)
West	4,450 (21.8)	3,349 (31.4)	2,997 (28.6)	3,240 (30.9)
Health plan type
HMO	3,786 (18.6)	1,693 (15.9)	1,692 (16.1)	1,677 (16.0)
PPO	15,432 (75.7)	8,478 (79.5)	8,316 (79.3)	8,307 (79.2)
CDHP or other	1,159 (5.7)	499 (4.7)	475 (4.5)	499 (4.8)
Insurance status
Fully insured	9,316 (45.7)	5,493 (51.5)	4,915 (46.9)	5,383 (51.4)
ASO	11,061 (54.3)	5,177 (48.5)	5,568 (53.1)	5,100 (48.7)
Coordination of benefits (> 0)	3,736 (18.3)	2,769 (26.0)	2,452 (23.4)	2,639 (25.2)
Medicare Advantage	663 (3.3)	513 (4.8)	480 (4.6)	476 (4.5)
Medicare Supplement	1,330 (6.5)	1,026 (9.6)	945 (9.0)	970 (9.3)
Deyo-Charlson Index comorbidities score, mean (SD)‡	2.5 (0.9)	2.5 (1.0)	2.5 (0.9)	2.5 (0.9)
Metastatic solid tumor	3,519 (17.3)	1,566 (14.7)	1,632 (15.6)	1,558 (14.9)
Neutropenia	608 (3.0)	91 (0.9)	78 (0.7)	91 (0.9)
Anemia	1,437 (7.1)	633 (5.9)	617 (5.9)	627 (6.0)

Abbreviations: ASO, administrative services only; CDHP, consumer-driven health plan; CVC, central venous catheter; HMO, health maintenance organization (includes point of service plan); IV, intravenous; PPO, preferred provider organization (includes fee-for-service plan); SD, standard deviation.

^*All differences between groups are significant at P < .001, except Deyo-Charlson Index comorbidities score (P = .44).

^†Matched groups are balanced for all covariates used for matching with standardized difference (stdiff) < 0.1, except age categories (stdiff = 0.18) , residence region (stdiff = 0.12), and trastuzumab use (stdiff = 0.24), which were not included in matching process.

^‡Revised by removing metastatic solid tumor.

Adjusting for demographic characteristics and treatment, we observed substantial variation in practice with regards to the types of venous access device used (Table 2). Patients who received trastuzumab and those with at least one claim indicating metastatic disease were more likely to have a long-term CVC than those without these characteristics (adjusted odds ratio [AOR] = 2.36 and 1.11; P < .001 and P = .01, respectively). Nevertheless, when controlling for differences in patient and treatment characteristics, there was significant variation in the use of long-term CVCs across regions, health plans, and funding sources. After adjusting for patient characteristics and treatment, the odds of long-term CVC insertion were 2.07 (95% CI, 1.93 to 2.23) in the South and 1.81 (95% CI, 1.68 to 1.94) in the Midwest compared with the West. Patients had higher odds of long-term CVC use if they lived in a rural area (AOR for rural v urban, 1.23; P < .001). Patients had lower odds of long-term CVC use if they were fully insured by the health plan versus administrative services only members, meaning the member's employer carried the insurance risk (AOR = 0.85; P < .001).

Table 2.

Long-Term CVC Insertion by Selected Patient Demographic and Treatment Characteristics

Characteristic	No. of Patients (N = 31,047)	Proportion of Patients With Long-Term CVC Insertion		Adjusted Odds Ratio*†	P
		Unadjusted (actual)	Adjusted*
Residence region
Northeast	6,725	57	23	0.90	.01
Midwest	7,607	72	37	1.81	< .001
South	8,916	75	41	2.07	< .001
West	7,799	57	25	Reference
Residence area type
Rural	5,311	71	33	1.23	< .001
Urban	24,112	64	29	Reference
Health plan type
HMO	5,480	69	34	1.18	.02
PPO	23,909	65	29	0.93	.23
CDHP or other	1,658	70	30	Reference
Insurance status
Fully insured	14,809	63	29	0.85	< .001
ASO	16,238	68	33	Reference
Trastuzumab user‡
Yes	6,696	79	41	2.36	< .001
No	24,351	62	23	Reference
Metastatic solid tumor
Yes	5,085	69	32	1.11	.01
No	25,962	65	30	Reference

Abbreviations: ASO, administrative services only; CDHP, consumer-driven health plan; COB, coordination of benefits; CVC, central venous catheter; HMO, health maintenance organization (includes point of service plan); PPO, preferred provider organization (includes fee-for-service plan).

^*Adjusted for age, index year, Medicare Advantage status, Medicare Supplement status, COB, Deyo-Charlson Index, metastatic solid tumor, neutropenia, anemia, myocardial infarction, congestive heart failure, peripheral vascular disease, cerebrovascular disease, dementia, chronic obstructive pulmonary disease, rheumatological disease, peptic ulcer disease, mild liver disease, diabetes (mild to moderate), hemiplegia or paraplegia, moderate or severe renal disease, diabetes and complications, malignancy, moderate to severe liver disease, AIDS, mastectomy (partial or total), lumpectomy, lymph node–directed surgery, radiation therapy or brachytherapy, complications, total health care cost, and number of injectable chemotherapies in preindex period.

^†Odds of a member having a long-term CVC in the group of interest compared with the reference group.

^‡Measured over the entire study period.

In unadjusted analysis of matched pairs, we observed a higher proportion of members with complications among those with long-term CVCs compared to those with temporary IVs (Table 3). Approximately 6% of women with a long-term CVC experienced a bloodstream infection, and 9% had a thrombovascular complication, compared with 2% and 4% of their matched peers with temporary IV catheters (both Ps < .001).

Table 3.

Complications Risk by Long-Term CVC Use, Matched Analysis

Complication	Temporary IV Catheter (n = 10,483)		Long-Term CVC (n = 10,483)		HR, Unadjusted Model (95% CI)‡	AHR, Adjusted Model (95% CI)§	Model Sensitivity Analysis∥ (range of AHRs obtained from models)§
	Pre-index Period, (% of total)*	Postindex Period, (% of total)†	Pre-index Period, (% of total)*	Postindex Period, (% of total)†
Bloodstream infections	0.5	2.1	0.6	5.6	2.68 (2.30 to 3.13)	2.70 (2.31 to 3.16)	2.66-2.73
Thrombovascular complications	3.1	3.9	3.3	9.3	2.48 (2.21 to 2.78)	2.61 (2.33 to 2.93)	2.59-2.60

NOTE. Hazards of experiencing postindex complication if member has a long-term CVC compared to a temporary IV catheter.

Abbreviations: AHR, adjusted hazard ratio; CVC, central venous catheter; HR, hazard ratio; IV, intravenous.

^*Pre-index period defined as 12 months before index long-term CVC insertion date, or index chemotherapy claim date for members with temporary IV catheter. Pre-index differences between long-term CVC and temporary IV catheter groups were not significant for any complication as a result of matching.

^†Postindex period defined as up to 12 months after index long-term CVC insertion date, or index chemotherapy claim date for members with temporary IV catheter. Postindex differences between long-term CVC and temporary IV catheter groups were significant for all complications at P < .001.

^‡Unadjusted model compares the postindex period proportion of patients with a complication in the long-term CVC group with the temporary IV catheter group.

^§Adjusted for age, index year, health plan region, Medicare Advantage status, Medicare Supplement status, insurance status, Deyo-Charlson Index, metastatic solid tumor, neutropenia, anemia, malignancy, moderate-to-severe liver disease, cerebrovascular disease, chronic obstructive pulmonary disease, mastectomy (partial or total), lumpectomy, lymph node–directed surgery, radiation therapy or brachytherapy, preindex complications, total health care cost, and number of injectable chemotherapies in preindex period.

^∥Alternative models include adjustment for health plan type, coordination of benefits, and median household income in zip code of residence, myocardial infarction, congestive heart failure, peripheral vascular disease, dementia, rheumatological disease, peptic ulcer disease, mild liver disease, diabetes (mild to moderate), hemiplegia or paraplegia, moderate or severe renal disease, diabetes and complications, AIDS, interaction terms of long-term CVC use by age, by metastatic solid tumor and by preindex complication (with AHR and 95% CI).

In adjusted matched-pairs analyses (Table 3), the risk of venous complications was more than two-fold higher for patients with long-term CVCs compared to those with IV catheters for bloodstream infections (adjusted hazard ratio, 2.70; 95% CI, 2.31 to 3.16) and thrombovascular complications (adjusted hazard ratio, 2.61; 95% CI, 2.33 to 2.93). These findings held for both outcomes when we tested alternative models that included health plan type, coordination of benefits, median household income in zip code of residence, and comorbid conditions. The inclusion of interaction terms did not change the interpretation of the main effect.

Discussion

In this analysis of commercially insured women treated with chemotherapy for breast cancer, we observed widespread variation in the use of long-term CVCs, as well as an increased risk of complications associated with their use. The risk was more than two-fold higher for both bloodstream infections and thrombovascular complications. Our findings were similar to those of a previous study of Medicare beneficiaries with breast cancer, age 66 years and older, despite differences in the study cohorts.¹⁵

Our findings here are consistent with those of other studies that estimated incidence of complication risk associated with long-term CVC use.^5–8 In another recent claims-based study of CVC use in 5,000 patients with cancer from a large health plan, the most frequent complications within 90 days of CVC insertion were bloodstream infections and thrombosis. Rates were 4.3 and 6.7 per 1,000 catheter days, respectively.⁵ Use of CVCs may partially explain the rising rates of thrombotic complications observed in patients with cancer.²³

We observed substantial variation in the use of long-term CVCs by region, health plan, and funding source. In the South, 41% of patients with breast cancer had a CVC inserted compared with approximately 25% of patients in the Northeast and West regions of the United States. After adjusting for patient characteristics and treatments, patients treated in the South had a two-fold increase in the use of CVCs, suggesting that some CVC use is discretionary, reflecting provider or institutional differences in practice patterns. Similar patterns were also observed across health plans and funding sources, which also suggests discretionary use. Given the morbidity associated with long-term CVCs, health plan and practice-level medical policies should take these risks into consideration in order to limit CVC use to situations with poor venous access or where the need for prolonged infusion therapy, such as trastuzumab, is required. Although the ASCO guidelines provide guidance on catheter type, insertion site, and placement, as well as prophylaxis and management of both catheter-related infection and thrombosis, patient selection for CVCs is not addressed.² Given the wide variation in practice patterns, evidence-based guidelines and tools to help identify patients for whom a long-term CVC is needed and those for whom a temporary IV catheter is likely to be sufficient are desirable.

The results of this study should be interpreted in the light of several limitations. First, differences in complications between groups could be explained by unmeasured confounders. For example, if factors such as stage of disease, socioeconomic status, treatment regimen, and patient preferences influenced both the use of long-term CVCs and the likelihood of complications, they could confound the association we observed between this exposure and outcome. Second, although claims data are extensively used in real-world epidemiology studies, coding may be inaccurate in some cases. For example, the use of ICD-9 codes may have mis-assigned members as having or as not having a new breast cancer diagnosis.^24,25 To reduce the chance of classifying a patient as having breast cancer when she did not, we required members to have at least two claims with a breast cancer diagnosis or a claim with a breast cancer diagnosis and procedure code for cancer-directed treatment, as well as intravenous chemotherapy. Codes for long-term CVC insertion have not been extensively validated for accuracy, but they were updated in 2004 in response to imprecise coding.^26,27 Codes for long-term CVC removal were not used reliably, limiting our ability to quantify the duration of long-term CVC exposure. Future research is needed to accurately determine the length of catheter exposure and associated risk by integrating claims and electronic medical record data. Last, the codes used to identify complications included in this study have not been extensively validated in the claims data. Therefore, some complications identified may be related to other, non-catheter sources. Because of this limitation, we compared complication risk between members with long-term CVCs and temporary IV catheters to account for non–catheter-related complications that would be expected to be similar between groups.

Our findings, based on the experience of women treated in routine care settings, suggest that long-term CVC use is associated with significant morbidity. Despite their advantages for patients who have poor venous access or need for prolonged infusion therapy, there may be opportunities to reduce the use of long-term CVCs under discretionary circumstances, and thus improve patient care. Evidence-based guidelines and tools may help decrease discretionary use of long-term CVCs. Future research is needed to develop additional strategies to reduce device-related complications when long-term CVCs are needed.

Appendix

Table A1.

Claims Codes for Procedures and Conditions of Interest

Injectable chemotherapy agents used for breast cancer	HCPCS codes: J9000-J9002, J9035, J9045, J9060, J9062, J9070, J9080, J9090-J9097, J9170, J9171, J9178, J9179, J9181, J9182, J9190, J9201, J9207, J9250, J9260, J9264, J9265, J9293, J9306, J9354, J9355, J9360, J9370, J9375, J9380, C9240, C9257, C9280, C9292, Q2024, Q2048-Q2050
Insertion of long-term CVCs	CPT codes: 36556, 36558, 36561, 36563, 36565, 36566
Catheter-related complications
Bloodstream infections	ICD-9 diagnosis codes: 038.0, 038.10, 038.11, 038.12, 038.19, 038.2, 038.3, 038.40, 038.41, 038.42, 038.43, 038.44, 038.49, 038.8, 038.9, 790.7, 995.91, 995.92, 996.62, 999.31, 999.39
Thrombovascular complications	ICD-9 diagnosis codes: 459.0, 459.10, 459.11, 459.12, 459.13, 459.19, 459.2, 459.30, 459.31, 459.32, 459.33, 459.39, 459.81, 459.89, 459.99, 998.11, 444.0, 444.1, 444.21, 444.22, 444.81, 444.89, 444.9, 451.0, 451.11, 451.19, 451.2, 451.81, 451.82, 451.83, 451.84, 451.89, 451.9, 453.0, 453.1, 453.2, 453.3, 453.40, 453.41, 453.42, 453.8, 453.9, 415.11, 415.12, 415.19, 433.00, 433.10, 433.20, 433.30, 433.80, 433.90, 434.00, 434.01, 434.10, 434.11, 434.90, 434.91, 435.8, 435.9, 436, 445.01, 445.02, 445.81, 445.89, 999.2
	ICD-9 procedure code: 99.10
	CPT codes: 34001, 34051, 34101, 34111, 34151, 34201, 34203, 34401, 34421, 34451, 34471, 34490, 37201, 33910, 33915, 33916, 37195, 75896, 92973, 92975, 92977

Abbreviations: CPT, Current Procedural Terminology; CVC, central venous catheter; HCPCS, Healthcare Common Procedure Coding System; ICD-9, International Classification of Diseases, ninth edition.

Authors' Disclosures of Potential Conflicts of Interest

Disclosures provided by the authors are available with this article at jop.ascopubs.org.

Author Contributions

Conception and design: Allison Lipitz-Snyderman, Michael F. Pollack, John Barron, Elena B. Elkin, Peter B. Bach, Jennifer L. Malin

Collection and assembly of data: Qinli Ma, Michael F. Pollack

Data analysis and interpretation: All authors

Manuscript writing: All authors

Final approval of manuscript: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Complications Associated With Use of Long-Term Central Venous Catheters Among Commercially Insured Women With Breast Cancer

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or jop.ascopubs.org/site/misc/ifc.xhtml.

Allison Lipitz-Snyderman

No relationship to disclose

Qinli Ma

Employment: HealthCore/Anthem

Michael F. Pollack

Employment: HealthCore/Anthem

Stock or Other Ownership: Anthem

John Barron

Employment: HealthCore/Anthem

Stock or Other Ownership: Anthem

Research Funding: Anthem, numerous pharmaceutical companies (Inst)

Elena B. Elkin

No relationship to disclose

Peter B. Bach

Leadership: Exam Works

Stock or Other Ownership: Exam Works

Honoraria: Genentech, Defined Health, American Journal of Managed Care

Consulting or Advisory Role: Foundation Medicine

Travel, Accommodations, Expenses: Genentech, Defined Health, American Journal of Managed Care

Jennifer L. Malin

Employment: Anthem

Stock or Other Ownership: Anthem