- Email: [email protected]
Central venous catheter infections in home parenteral nutrition patients: Outcomes from Sustain: American Society for Parenteral and Enteral Nutrition's National Patient Registry for Nutrition Care
American Journal of Infection Control 44 (2016) 1462-8
Contents lists available at ScienceDirect
American Journal of Infection Control
American Journal of Infection Control
j o u r n a l h o m e p a g e : w w w. a j i c j o u r n a l . o r g
Major Article
Central venous catheter infections in home parenteral nutrition patients: Outcomes from Sustain: American Society for Parenteral and Enteral Nutrition’s National Patient Registry for Nutrition Care Vicki M. Ross PhD, RN, CNSC a, Peggi Guenter PhD, RN, FAAN a,*, Mandy L. Corrigan MPH, RD, CNSC, FAND a, Debra Kovacevich MPH, RN b, Marion F. Winkler PhD, RD, LDN, CNSC, FASPEN c, Helaine E. Resnick PhD, MPH d, Tina L. Norris PhD d, Lawrence Robinson PharmD, MS e, Ezra Steiger MD, FACS, FASPEN, AGAF f a
American Society for Parenteral and Enteral Nutrition, Silver Spring, MD University of Michigan Home Care Services, University of Michigan, Ann Arbor, MI c Rhode Island Hospital and Alpert Medical School of Brown University, Providence, RI d Silver Spring, MD e Collierville, TN f Digestive Disease Institute, Cleveland Clinic, Cleveland, OH b
Key Words: Database CLABSI
Background: Home parenteral nutrition (HPN) is a high-cost, complex nutrition support therapy that requires the use of central venous catheters. Central line-associated bloodstream infections (CLABSIs) are among the most serious risks of this therapy. Sustain: American Society for Parenteral and Enteral Nutrition’s National Patient Registry for Nutrition Care (Sustain registry) provides the most current and comprehensive data for studying CLABSI among a national cohort of HPN patients in the United States. This is the first Sustain registry report detailing longitudinal data on CLABSI among HPN patients. Objective: To describe CLABSI rates for HPN patients followed in the Sustain registry from 2011-2014. Methods: Descriptive, χ2, and t tests were used to analyze data from the Sustain registry. Results: Of the 1,046 HPN patients from 29 sites across the United States, 112 (10.7%) experienced 194 CLABSI events during 223,493 days of HPN exposure, for an overall CLABSI rate of 0.87 episodes/1,000 parenteral nutrition-days. Although the majority of patients were female (59%), adult (87%), white (75%), and with private insurance or Medicare (69%), CLABSI episodes per 1,000 parenteral nutrition-days were higher for men (0.69 vs 0.38), children (1.17 vs 0.35), blacks (0.91 vs 0.41), and Medicaid recipients (1.0 vs 0.38 or 0.39). Patients with implanted ports or double-lumen catheters also had more CLABSIs than those with peripherally inserted or central catheters or single-lumen catheters. Staphylococci were the most commonly reported pathogens. These data support findings of smaller studies about CLABSI risk for children and by catheter type and identify new potential risk factors, including gender, race, and insurance type. Conclusions: Additional studies are needed to determine effective interventions that will reduce HPN-associated CLABSI. © 2016 Association for Professionals in Infection Control and Epidemiology, Inc. Published by Elsevier Inc. All rights reserved.
* Address correspondence to Peggi Guenter, PhD, RN, FAAN, American Society for Parenteral and Enteral Nutrition, 8630 Fenton St. Suite 412 Silver Spring, MD 20910. E-mail address: [email protected] (P. Guenter). Sustain was supported, in part, by unrestricted scientific grants from Baxter Healthcare Corporation, the C. R. Bard Foundation, and the A.S.P.E.N. Rhoads Research Foundation. Conflicts of interest: None to report.
Assessments of high-risk, complex patient populations require large datasets or registries to generate detailed outcome data. Registries are especially useful to describe a real-world picture of patient populations to better understand their disease course.1 Recognizing the need for data that accurately describe the characteristics and outcomes of high-risk, complex patients receiving parenteral nutrition (PN) at home, the American Society for Parenteral and Enteral Nutrition (A.S.P.E.N.) launched Sustain: A.S.P.E.N.’s National Patient
0196-6553/© 2016 Association for Professionals in Infection Control and Epidemiology, Inc. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ajic.2016.06.028
V.M. Ross et al. / American Journal of Infection Control 44 (2016) 1462-8
Registry for Nutrition Care (Sustain registry) in 2011. The Sustain registry was designed to collect and compile data on adult and pediatric home parenteral nutrition (HPN) patients.2 The Sustain registry generates longitudinal observational data that can be used to describe and understand the HPN population in the United States and provide evidence for future research. Developed initially to provide intravenous nutrition for acutely ill hospitalized patients with intestinal failure, PN therapy became an option for home care in 1969. HPN allowed patients to receive this complex nutrition therapy in a familiar, comfortable environment. Along with the convenience of administering PN at home came the significant responsibility for patients and their family caregivers to provide direct care.3 Occurring both in the hospital and at home, central lineassociated bloodstream infections (CLABSIs) are among the most common and serious complications of PN therapy. For hospitalized patients there are standardized definitions and data collection methods to define and determine CLABSI that in turn allow for benchmarking infection rates, research activities, and reviews by the public and third-party payers. According to a 2013 National Healthcare and Safety Network Report, 4 CLABSI rates for hospitalized patients in nonburn units range from 0-1.2 per 1,000 catheterdays. Remarkably, there are no corresponding standardized definitions or mandatory data reporting for HPN patients. The number of HPN patients in the United States is currently unknown. A 2010 survey from the National Home Infusion Association estimated that 33,000 people in the United States require HPN. The same survey estimated that annual spending for HPN was $652 million.5 These estimates are likely lower than the actual numbers of HPN patients because not all HPN providers participated in the National Home Infusion Association survey. In the United States, there are few large datasets describing characteristics and clinical outcomes of HPN patients. Historical data from 19851992 are available from the Oley-A.S.P.E.N. Information System and the North American Home Parenteral and Enteral Patient Registry has some data from 1992-1994.6 In 2005, a home care company published outcomes from a 5-year retrospective evaluation of their experience with 4,540 HPN patients.7 These databases are no longer current or available for public use, leaving significant gaps in knowledge concerning outcomes of HPN patients in the United States. HPN rates of infections reported in a systematic review of published studies ranged from 0.38-4.58 episodes per 1,000 catheterdays. The lack of standardized definitions, data collection methods, and reporting mechanisms for HPN patients contribute to the wide range in reported rates.8 There is an obvious need for current data on CLABSI rates in HPN patients. The Sustain registry database provides the most comprehensive and current HPN patient information in the United States. The purpose of this longitudinal study was to describe the CLABSI rate and characteristics for HPN patients followed in the Sustain registry from 2011-2014. METHODS Nutrition support professionals who provided HPN services in outpatient or home-care sites were invited to participate in the Sustain registry by providing de-identified observational or chart data about their HPN patients. To ensure protection of human subjects, A.S.P.E.N. received approval for the overall conduct of the Sustain registry from the New England Institutional Review Board and participant providers at individual sites obtained requisite approvals from their local institutional review boards. As required by their local institutional review boards, participating providers obtained written consent or waivers of consent from their HPN patients. Additional details about the development and conduct of the Sustain registry are provided in earlier publications.2,9 The registry system has a built-in quality control
1463
program to alert site participants and registry administration if duplicate patients are entered. In the few instances that this occurred, the registry administrator worked out this issue with the sites. Baseline observations from chart data collection included demographic, HPN, and catheter characteristics. Follow-up observations from chart data included information about occurrence of CLABSIs and pathogens associated with these events. When CLABSIs were reported, participating sites were directed to provide information about the pathogen associated with the infection. HPN duration was defined as the time in days HPN patients were followed in the Sustain registry from baseline to the final followup visit or patient contact. Because Sustain is an observation type registry, the time frames between the baseline and 1 or more followup visits were determined by participating providers based on their clinical process. In most cases, a follow-up encounter corresponded with an episode of care (eg, outpatient visit, hospitalization, and routine visit), but some encounters involved telephone contact. This study reports CLABSI data for patients who were enrolled in the Sustain registry between August 2011 and February 2014 and who had at least 1 follow-up visit or contact through August 2014. Definitions Patients starting HPN therapy within 90 days of baseline data collection were defined as “new,” and their PN-days were calculated from the actual HPN start date. Patients receiving HPN therapy for >90 days before baseline data collection were defined as “existing,” and their PN-days were calculated from the date of baseline enrollment in the registry. No attempt was made to ascertain the exact date of HPN initiation for the “existing” patients. Thus, PNdays and CLABSI rates reflect the time patients were receiving HPN and followed in the Sustain registry. CLABSIs were defined as primary bloodstream infections in patients who had a central venous catheter (CVC) within the 48hour period before the development of the bloodstream infection and not related to infections at other body sites. This definition was shared with each site during Sustain registry training and placed on the help pages in the online data collection forms. Each participating site determined whether a patient’s infection met the criteria for CLABSI. The overall CLABSI rate was calculated as the number of CLABSI events per 1,000 PN-days. Patient-level characteristics were also assessed in terms of presence or absence of at least 1 CLABSI, and these rates were calculated on the patient level. Data analysis Data analyses included the 1,046 patients who were enrolled in the Sustain registry between August 2011 and February 2014, and who had at least 1 follow-up visit or contact through August 2014. Overall CLABSI rates were calculated as events per unit time and corresponding rates for demographic and clinical characteristics were calculated on the patient level among those with 1 or more CLABSI episodes. To determine whether CLABSI status differed by demographic, clinical, or catheter-related characteristics, χ2 tests for categorical and t tests for continuous data were used. Analyses were conducted using SAS (SAS Institute Inc, Cary, NC). RESULTS Site characteristics Twenty-one of 29 participating sites were hospitals and the remaining sites were home infusion companies. Of the hospital-based sites, 35% were academic medical centers with adult patients alone or adult and pediatric patients, 60% were free-standing academic
1464
V.M. Ross et al. / American Journal of Infection Control 44 (2016) 1462-8
children’s hospitals, and 5% were community hospitals. Hospitals were located in 26 cities across 15 states. The size of hospital-based sites varied from 430-1,000 beds in academic medical centers, and 85427 beds in children’s hospitals. Sixty-nine percent of HPN patients, of whom the majority were adults, were enrolled by medical centers and hospitals, 19% were from home care or infusion service providers with both adult and pediatric patients, and the remaining 11% were enrolled at children’s hospitals. Half of the sites enrolled all “new” HPN patients and 60% approached all “existing” HPN patients to participate. Most sites (83.3%) consecutively enrolled and followed HPN patients during their participation in the Sustain registry. Sample characteristics Of all 1,251 Sustain registry patients with the baseline data previously described,9 1,046 had at least 1 follow-up visit or contact with their participating site provider and met eligibility criteria for this longitudinal, descriptive study. In this observational study, although not common it was possible for patients to be entered into the registry and return 1 day later for their last recorded follow-up visit. Mean HPN duration was 214 days (range, 1-946 days). Adult patients received HPN an average of 202 days, and pediatric patients received HPN an average of 291 days. There were 194 CLABSI events in 112 patients. The overall CLABSI rate was 0.87 episodes per 1,000 PN-days. Demographic characteristics Table 1 shows the demographic characteristics of the sample by CLABSI status. A common profile of a Sustain registry patient was an adult, white, woman. However, when comparing those who had at least 1 CLABSI to those with no infections for different demographic categories, children, nonwhites, and males accounted for 13.2% of the overall sample, but 42% of those experienced 1 or more CLABSI and only 9.7% had no CLABSI. Similarly, blacks accounted for 12% of the sample but 20.5% of those who had at least 1 CLABSI. These differences were also reflected in the higher infection rates for children, blacks, and males.
Among patients with data about insurance coverage, most had private or commercial insurance, or Medicare, followed by Medicaid alone or in combination. Insurance coverage differed by infection status. Nearly 47% of patients with at least 1 infection reported Medicaid coverage. Patients receiving Medicaid had CLABSI rates that were more than twice the rate of those with Medicare or private pay (1.0 vs 0.38 or 0.39/1,000 HPN-days). Clinical characteristics The 5 most common diagnoses reflected some degree of bowel dysfunction, with CLABSI occurring significantly more frequently in the short bowel group (Table 2). Unexpectedly, the 18.7% reporting gastrointestinal cancer diagnosis accounted for <5% of patients with CLABSI. Data about the presence or absence of small and large bowel stomas were collected for 299 patients with short bowel syndrome. Of these patients, 57.3% had a small bowel stoma and 8.4% had a large bowel stoma. There was no statistically significant difference in the number of infections between patients with or without stomas. More than half (60.8%) of patients in the sample had HPN duration of 6 months or less and 21% were followed for a year or more while receiving HPN. CLABSIs were significantly more frequent in the patients categorized as longer HPN duration; yet, when accounting for the number of days of exposure, the CLABSI rates per 1,000 PN-days were lowest for those followed a year or longer. More than 90% of patients in this study received daily infusions of HPN and 35.2% received intravenous lipid emulsions daily. There were no statistically significant differences for the occurrence of CLABSI when comparing patients receiving daily HPN to those receiving HPN fewer times per week or for patients receiving daily lipids to those receiving lipids fewer times per week. Catheter characteristics Ninety percent of patients administered their HPN through a tunneled catheter (47%) or a peripherally inserted central catheter (PICC)
Table 1 Selected demographic characteristics by infection status at follow-up, Sustain: American Society for Parenteral and Enteral Nutrition’s National Patient Registry for Nutrition Care
Variable Total Gender Women Men Age group (y) < 18 ≥ 18 Race White Black Other Ethnicity Hispanic Non-Hispanic Insurance Medicare Medicaid Private Personal pay Medicare supplement Other
Entire sample n (%) 1,046 (100) 1,045 (100) 614 (58.8) 431 (41.2) 1,046 (100) 138 (13.2) 908 (86.8) 1,046 (100) 780 (74.6) 125 (12) 141 (13.5) 927 (100) 63 (6.0) 864 (93.2) 982 (100) 224 (22.8) 196 (20.0) 500 (50.9) 15 (1.5) 22 (2.2) 25 (2.5)
Total parenteral nutrition-days
CLABSI rate* ‡
223,493
0.87
136,112 87,200
0.38 0.69
40,133 183,360
1.17 0.35
171,528 25,217 26,748 175,505 13,270 162,235
0.41 0.91 0.71 0.90 0.51
44,669 48,016 80,305 1,634 5,280 3,107
0.38 1.00 0.39 0.61 0.19 1.61
Patients with 1 or more CLABSI n (%)
Patients with no CLABSI n (%)
112 (10.7) 112 (100) 52 (46.2) 60 (53.6) 112 (100) 47 (42) 65 (58) 112 (100) 70 (62.5) 23 (20.5) 19 (7.0) 95 (100) 12 (12.6) 83 (87.4) 103 (100) 17 (16.5) 48 (46.6) 31 (30.1) 1 (1.0) 1 (1.0) 5 (4.8)
934 (89.3) 933 (100) 562 (60.2) 371 (39.8) 934 (100) 91 (9.7) 843 (90.3) 934 (100) 710 (76.0) 102 (10.9) 122 (13.1) 832 (100) 51 (5.5) 781 (93.9) 879 (100) 207 (23.5) 148 (16.8) 469 (53.4) 14 (1.6) 21 (2.4) 20 (2.3)
CLABSI, central line-associated bloodstream infection. *CLABSI rates are presented at the patient level unless otherwise indicated. †P values reflect χ2 or Fisher exact test, as appropriate. ‡ Overall CLABSI rate for the sample is based on total number of CLABSI events divided by total parenteral nutrition-days.
P value† .0050
< .0001
.0035
.0171
.0886 < .0001 < .0001 1.0000 .5004 .1766
V.M. Ross et al. / American Journal of Infection Control 44 (2016) 1462-8
1465
Table 2 Selected clinical characteristics by infection status at follow-up, Sustain: American Society for Parenteral and Enteral Nutrition’s National Patient Registry for Nutrition Care Entire sample n (%) Total Top-5 diagnoses Short bowel syndrome Gastrointestinal cancer Crohn’s disease Gastromotility/pseudo-obstruction disease Gastrointestinal bypass for obesity Duration of parenteral nutrition at follow-up (d) ≤ 90 91-180 181-365 > 365
Total parenteral nutrition-days
CLABSI rate*
Patients with 1 or more CLABSI n (%)
Patients with no CLABSI n (%)
1,046 (100)
223,493
0.87‡
112 (10.7)
934 (89.3)
299 (28.6) 196 (18.7) 171 (16.3) 116 (11.1) 75 (7.2)
108,378 25,823 48,727 35,077 10,120
0.48 0.19 0.25 0.48 0.1
52 (46.2) 5 (4.5) 12 (10.7) 17 (15.2) 1 (0.9)
247 (26.4) 191 (20.4) 159 (17) 99 (10.6) 74 (7.9)
485 (46.4%) 172 (16.4%) 166 (15.9%) 223 (21.3%)
20,916 22,328 42,901 137,348
0.67 0.67 0.68 0.39
14 (12.5%) 15 (13.4%) 29 (25.9%) 54 (48.2%)
P value†
< .0001 < .0001 .088 .1448 .0064 < .0001
471 (50.4%) 157 (16.8%) 137 (14.7%) 169 (18.1%)
CLABSI, central line-associated bloodstream infection. *CLABSI rates are presented on the patient level unless otherwise indicated. †P values reflect χ2 or Fisher exact test, as appropriate. ‡ Overall CLABSI rate for the sample is based on total number of CLABSI events divided by total parenteral nutrition-days.
Table 3 Catheter type, related care, and pathogens by infection status at follow-up, Sustain: American Society for Parenteral and Enteral Nutrition’s National Patient Registry for Nutrition Care
Variable Total Catheter type Peripherally inserted central catheter Port Tunneled Other Number of lumens Single Double Triple Use of ethanol lock No Yes Use of anti-infectives No Yes Dressing changed by: Patient and caregiver Caregiver Patient Visiting nurse
Entire sample n (%)
Total parenteral nutrition-days
CLABSI rate*
Patients with 1 or more CLABSI n (%)
Patients with no CLABSI n (%)
1,046 (100) N = 1,045 449 (42.9) 93 (8.9) 492 (47) 11 (1.1)
223,493
0.87‡
112 (10.7)
934 (89.3)
53,978 18,131 150,087 1,297
0.41 0.66 0.51 0.77
22 (19.6) 12 (10.7) 77 (68.8) 1 (0.9)
427 (45.7) 81 (8.7) 415 (44.4) 10 (1.1)
481 (46) 538 (51.4) 27 (2.6)
73,749 147,963 1,781
0.38 0.57 0.00
28 (25.0) 84 (75.0) 0 (0.0)
453 (48.5) 454 (48.6) 27 (2.9)
762 (72.8) 284 (27.2)
142,298 81,195
0.48 0.54
68 (60.7) 44 (39.3)
694 (74.3) 240 (25.7)
788 (75.3) 258 (24.7) 568 (100) 15 (2.6) 102 (18.0) 63 (11.1) 388 (68.3)
174,423 49,070
0.50 0.51
1,978 34,631 32,844 60,018
0.00 0.75 0.27 0.50
87 (77.7) 25 (22.3) 65 (11.4) 0 (0.0) 26 (40) 9 (13.8) 30 (46.2)
701 (75.1) 233 (24.9) 503 (88.6) 15 (3.0) 76 (15.1) 54 (10.7) 358 (72.2)
P value† < .0001
< .0001
.0022
.5425
< .0001†
CLABSI, central line-associated bloodstream infection. *CLABSI rates are presented on the patient level unless otherwise indicated. †P values reflect χ2 or Fisher exact test, as appropriate. ‡Overall CLABSI rate for the sample is based on total number of CLABSI events divided by total parenteral nutrition-days.
(42.9%). The remaining 9% reported using subcutaneous infusion ports. Table 3 contains data about catheter types and related care. Among patients with at least 1 CLABSI, 19.6% used a PICC compared with 68.8% of those with a tunneled catheter. Ninety-seven percent of patients had either a single- or doublelumen CVC. The CLABSI events were statistically more for the doublelumen CVC compared with the single-lumen CVC. There were no CLABSIs reported for the 27 patients with triple-lumen catheters. The use of ethanol locks were reported in 27.2% of patients who accounted for 39.3% of the infections, whereas the remaining 72.8% of the group did not use ethanol locks and accounted for 60.7% of the infections. Data from the medication inventory revealed that 24.7% of patients received anti-infective agents (oral or intravenous, antibiotics, or antifungal agents) at baseline. There was no statistically significant difference (P = .54) for CLABSI events between patients who did and did not report use of anti-infective agents. Information about who was changing CVC dressings was available for 568 patients. Data were not reported or missing for the
remaining patients. CLABSI rates were highest among patients for whom caregivers changed the dressings alone (0.75 events per 1,000 PN-days). There were no CLABSIs reported for the 15 HPN patients who reported dressings changed by patients and a caregiver. Identified pathogens There were 194 CLABSI episodes in 112 patients, and 96 patients had the pathogens identified. Reported pathogens included gram-positive, gram-negative, and fungal types. Sixteen percent of pathogens were listed as “other” without further detail. Figure 1 provides additional detail about the specific pathogens reported. DISCUSSION AND LIMITATIONS Nearly 11% of patients in our study experienced one or more CLABSI, supporting the concern that CLABSIs continue to remain for HPN patients. Furthermore, the most vulnerable and potentially
1466
V.M. Ross et al. / American Journal of Infection Control 44 (2016) 1462-8
Fig 1. Central line-associated bloodstream infection pathogens.
disenfranchised patients—children, minorities, and those receiving public assistance—experienced significantly higher numbers of infections. Our findings support the results of smaller or singlesite studies and provide additional evidence about the need to improve HPN care. Smaller studies with a variety of defining criteria for HPNrelated infections contribute to the wide range of reported infections in the literature.8 Although the exact number of HPN patients in the United States is not known, the overall CLABSI at the patient level, the rate for patients in the Sustain registry (0.87 per 1,000 PN-days; pediatric rate is 1.17 per 1,000 catheter-days and the adult rate is 0.35 per 1,000 PN-days) is within the range of other HPN noncancer pediatric studies of 0.19 and 2.14 per 1,000 catheter-days.8 Similarly, our CLABSI rate of 1.17 per 1,000 HPN-days is within the range reported by others (0.8010 and 9.811 infections per 1,000 catheter-days). Black and Hispanic patients in the Sustain registry experienced more than twice the rate of infection compared with whites. This is among of the first studies to highlight the potential risk, either directly or indirectly, that race has on HPN risk. Nguyen et al12 in a study of inpatients with inflammatory bowel disease, a condition associated with PN use, found malnourished African Americans were less likely than whites to receive PN (odds ratio, 0.67; 95% confidence interval, 0.50-0.89) and Hispanics had higher rates of catheter-related complications (5.1% vs 12.2%; P = .04). Whether race is directly or indirectly related, this new information underscores the importance of determining why these differences occur to be able to develop effective interventions and culturally appropriate educational materials. In our study, Medicaid patients had more than twice the rate of infection as those with private insurance. In at least 1 other study, the CLABSI rates were higher for pediatric patients with Medicaid insurance.13 More research is needed to determine and address the specific factors related to these findings. It is possible that Medicaid is a proxy for socioeconomic status. Alternately, it may reflect patient age, with higher infection rates among children who also use Medicaid. Notably, the CLABSI rates for Medicare and private insurance (0.38 vs 0.39/1,000 HPN-days) in our study were almost identical, suggesting chronological age of adults may have minimal influence on infection risk. The specific differences in CLABSI rates
by insurance coverage cannot be determined by registry data. These findings support the need for further study. HPN is considered a long-term therapy requiring a long-term catheter. Almost half of patients in the Sustain registry were using a PICC, with a higher number of CLABSIs in the implanted or and tunneled catheter groups. The proportion of PICCs reported in our study align with a study of trends using the Canadian HPN patient registry, where there was an increased use of PICCs from the initial 21.6% during the 2005-2008 period to 54.9% during the 20112014 period.14 There are at least 2 HPN studies suggesting PICCs are not associated with a higher risk of infection when compared with other CVCs. In a 5-year prospective study of HPN patients there were no infections reported for the PICC group and 44 and 20 per 1,000 catheter-days for the implanted and tunneled catheter groups, respectively.15 In another prospective study of HPN cancer patients, there were no infections reported for the 65 patients with a PICC and infection rates of 0.97 and 0.45 per HPN-days for the tunneled and implanted catheter groups, respectively.15 In a metaanalysis of 23 studies and more than 57,000 patients requiring CVCs for varied purposes, PICCs were less likely to be associated with CLABSI than other CVCs.16 Patients in the Sustain registry with double-lumen catheters experienced more CLABSIs than those with single-lumen catheters. These findings support recommendations to use a single-lumen CVC or the fewest number of lumens necessary to manage patient care.17,18 Surprisingly, there were no CLABSIs reported for the 27 Sustain registry patients using triple-lumen catheters. The patients in this group had a shorter average follow-up time compared with the doublelumen group (66 days vs 275 days). Further research is needed to determine other factors in addition to the number of catheter lumens that increase risk for CLABSI. For the 194 CLABSI episodes in 112 patients, 96 patients had pathogens identified. The majority of pathogens could be categorized as gram-positive, gram-negative, or fungal species. However, of the patients with identified pathogens, 16% were listed as “other” with no more specific identification. Our findings are similar to the pattern described by others who found the majority of infections were gram-positive infections followed by gram-negative and fungal infections.8,13 Registry data did not contain detailed information about
V.M. Ross et al. / American Journal of Infection Control 44 (2016) 1462-8
the antibiotic susceptibility of the infecting organisms; however, the general pattern of pathogens reinforces the need for research that focuses on the prevention or eradication of these specific pathogens. Furthermore, given the number of patients in our study reporting anti-infective therapy at baseline, more detailed studies about antibiotic resistance and potentially changing patterns of pathogens are needed. The primary diagnoses of Sustain registry patients reflected some degree of intestinal impairment. Surprising findings from this study were the reported CLABSI events for patients with cancer, stomas, and ethanol lock therapy. The low CLABSI rate for cancer patients in the Sustain registry was somewhat surprising given the higher infection rates of 0.54-6.8 per 1,000 catheter-days reported for HPN cancer patients in other studies.8 Only 5 of 196 patients in the Sustain registry with gastrointestinal cancer as the primary diagnosis experienced a CLABSI. It was outside the purpose of the Sustain registry to determine the exact type of cancer or cancer treatment. Our results were similar to the results of a prospective study of 254 European HPN cancer patients, the majority with solid tumor malignancies, reporting an overall CLABSI rate of 0.35 per 1,000 catheter-days.19 Because of the close proximity of catheters exiting the body and the enteric flora found at stoma sites, there is concern for crosscontamination and increased risk for infections when HPN patients have abdominal ostomies. In a 6-year observational cohort of 136 adult HPN patients, those with stomas and tunneled catheters experienced a decreased time to first CLABSI.20 Drews et al21 also found ostomies to be a risk factor for pediatric patients. Counter to the findings of others and similar to findings in the Sustain registry, Buchman et al10 found similar catheter-related infection rates for HPN patients with and without ostomies (0.34 vs 0.4 infections per 1,000-days). A more detailed study of patients with ostomies is needed to determine whether there are coexisting characteristics for some patients with ostomies that increase or ameliorate their risk of infection. For patients with recurrent infections, the Centers for Disease Control and Prevention recommends use of prophylactic antimicrobial lock solutions, of which ethanol is an example.17 More than a quarter of patients upon entry into the Sustain registry reported using ethanol locks, suggesting at least those using the lock solution may have experienced 1 or more CLABSI before enrollment. CLABSI events were significantly higher for patients with ethanol locks while followed in the Sustain registry compared with those who did not report ethanol lock at baseline. There is a paucity of data from prospective, randomized clinical trials about the most effective use of ethanol lock (eg, ideal catheter dwell time or frequency, strength, and dose).17,22 A detailed description about the use of ethanol locks was not a part of the Sustain registry and more research is needed to understand the use and effects of ethanol locks. Compared against their previous HPN experience, the ethanol group may have had fewer CLABSIs but continued to experience some CLABSIs while in the registry. Additionally, there was a national shortage of ethanol that occurred during the Sustain registry data collection and patients who were initially receiving the therapy may have been forced to reduce or eliminate the dose or frequency of the ethanol. Findings of this study are limited by the recruitment of volunteer sites and the use of data typically recorded by participating providers. The actual number of HPN patients is not known and it is difficult to determine how closely the volunteer sites approximated the actual HPN population or the bias that might have been introduced. Multiple attempts were made to communicate with and invite all HPN providers to participate in the Sustain registry. Although the Sustain registry did not have 100% participation, participants were located throughout the major regions of the United States and from both large and small sites, providing a more robust sample than what could otherwise be obtained from a single site or provider.
1467
Alternate explanations for some of our findings may be related to the noncompulsory recruitment or the detail of data routinely collected by participant providers and may not reflect the total HPN population. The purpose of the Sustain registry is to collect longitudinal, observational data about HPN patients that can be used in the development of future intervention testing with the ultimate goal of improving HPN care. Except to provide benchmarking data to the individual sites as a benefit of participating, there was no intent to evaluate the specific care provided by individual sites. The design of the Sustain registry does not allow for randomization or the testing of specific interventions. There were multiple checks within the Sustain registry system to avoid the same patient data from being entered twice by different providers and to address out-of-range or missing data. These checks are described in earlier publications.2,9 Sustain registry data are limited to information routinely collected by providers. Few data were missing for demographic variables such as age or gender, for example, but for data not routinely observed such as the person or persons changing dressings, a large portion of the data were missing. For some data that did not fall neatly into discrete categories such as multiple insurance types or anti-infective agents, recategorizing into nominal levels of yes or no most likely introduced some error in the analyses. Modeling data using a number of predictor variables as a function of HPN-days was discussed and determined not to be appropriate for this study for a couple of reasons. First, the purpose of our study was to provide descriptive data for future intervention testing research, so we opted to provide unadjusted rates without making statistical inferences. Second, the heterogeneity of the HPN population coupled with missing data for information that was not routinely collected by providers would result in an increasing number of cases lost as the number of predictor variables was added. For example, the number of CLABSI events for the pediatric population was significantly different from the adult population, and the unadjusted CLABSI rates were 3 times those of the adult population. To adjust rates based on modeling of predictor variables would have required a theoretical framework for selecting the appropriate predictor variables and a sample size that was large enough with complete data to provide adequate power. This study reflects HPN patients enrolled in the Sustain registry from 2011-2014. Trends in patient characteristics and conditions may change over time, and those changes may influence the data and study findings. For example, the number of PICCs used by HPN patients enrolled in the Sustain registry were surprising, but may reflect a trend found in the Canadian HPN registry.14 There is no universal health insurance plan in the United States and insurance plans even within the same carrier may differ, resulting in variable experiences for HPN patients. Implementation of the Affordable Care Act, which requires all US citizens to have insurance, began January 1, 2014.23 Data collection for this study began almost 3 years before and was almost complete before implementation of the Affordable Care Act provision. The influence of this provision on the health of HPN patients has yet to be determined, but the dramatic difference in CLABSI rates between those with and those without private or Medicare insurance underscores the need for further study. Conclusions The Sustain registry provides the most current and comprehensive data for HPN patients in the United States. Generalizability of the results are strengthened by standardized data collection methods and a large sample from providers across the United States. The findings from this study indicate HPN-associated CLABSIs are a continuing threat, especially to the most vulnerable populations (ie, young, minorities, and the poor). As in acute care, there is need for
1468
V.M. Ross et al. / American Journal of Infection Control 44 (2016) 1462-8
evidence-based interventions that will drive the CLABSI rate for HPN patients to zero. Unlike acute-care settings with hospital-based protocols used by trained health care personnel, safe and effective HPN therapy relies on patients and their family caregivers who must be able to receive, understand, and implement evidence-based information about HPN care. It is imperative that nutrition support teams, in particular physicians and nurses, convey and ensure implementation of best practices for insertion and care of CVCs when working with HPN patients and their families. Future research should examine the effects of sterile catheter insertion techniques, providing relevant educational information to HPN patients and their family caregivers about aseptic care. Continually providing feedback and support can provide the framework for lowering infection rates. Acknowledgment The authors thank the Sustain registry coordinator Katy Hanley and the data collection contributions of the following clinical sites: Arnold Palmer Hospital for Children (Orlando, FL), CarePoint Partners (Dallas, TX; Providence, RI; and Tampa, FL), Boston Children’s Hospital (Boston, MA), Children’s Hospital of Michigan (Detroit, MI), Children’s Mercy Hospital (Kansas City, MO), Children’s Hospital of Orange County (Orange, CA), Cleveland Clinic (Cleveland, OH), University of Pennsylvania Health System Clinical Nutrition Support Service (Philadelphia, PA), Connecticut Children’s Medical Center (Hartford, CT), Cook Children’s Medical Center (Fort Worth, TX), Equinox Healthcare Inc (Ellicott, MD), Geisinger Medical Center (Danville, PA), Indiana University Health Home Care (Indianapolis, IN), Intermountain Medical Center (Murray, UT), Le Bonheur Children’s Hospital (Memphis, TN), Nation’s Home Infusion LLC (Owings Mills, MD), Nationwide Children’s Hospital (Columbus, OH), Pediatric Home Service (Roseville, MN), Rady Children’s Hospital (San Diego, CA), Rhode Island Hospital and Hasbro Children’s Hospital (Providence, RI), Riley Hospital for Children at Indiana University Health (Indianapolis, IN), Sharp Home Infusion and Central Pharmacy Services (San Diego, CA), Spectrum Health Butterworth (Grand Rapids, MI), Texas Children’s Hospital Baylor College of Medicine (Houston, TX), ThriveRx (Cincinnati, OH), University of Michigan (Ann Arbor, MI), and University of Virginia Health System (Charlottesville, VA). References 1. Gliklich R, Dreyer N, Leavy M, editors. Registries for evaluating patient outcomes: a user’s guide. 3rd ed. AHRQ Publication No. 13(14)-EHC111. Two volumes. Rockville (MD): Agency for Healthcare Research and Quality; 2010. Available from: http://www.effectivehealthcare.ahrq.gov/registries-guide-3.cfm. Prepared by the Outcome DEcIDE Center [Outcome Sciences, Inc., a Quintiles company] under Contract No. 290 2005 00351 TO7. 2. Guenter P, Robinson L, DiMaria-Ghalili RA, Lyman B, Steiger E, Winkler MF. Development of SustainTM: A.S.P.E.N.’s National Patient Registry for Nutrition Care. JPEN J Parenter Enteral Nutr 2012;36:399-406.
3. Ayers P, Guenter P, Holcombe B, Plogsted S. Home parenteral nutrition. In: A.S.P.E.N. parenteral nutrition handbook. 2nd ed. Silver Spring (MD): American Society for Parenteral and Enteral Nutrition; 2014:247-85 [chapter 10]. 4. Dudeck MA, Edwards JR, Allen-Bridson K, Gross C, Malpiedi PJ, Peterson KD, et al. National healthcare safety network report, data summary 2013, device associated module. Am J Infect Control 2015;43:206-21. 5. National Home Infusion Association (NHIA). Phase I: 2010 NHIA provider survey comprehensive aggregate analysis report. Alexandria (VA): National Home Infusion Association; 2011. 6. Howard L. Home parenteral nutrition: survival, cost, and quality of life. Gastroenterology 2006;130(Suppl 1):S52-9. 7. Ireton-Jones C, DeLegge M. Home parenteral nutrition registry: a five-year retrospective evaluation of outcomes of patients receiving home parenteral nutrition support. Nutrition 2005;21:156-60. 8. Dreesen M, Foulon V, Spriet I, Goossens GA, Hiele M, De Pourcq L, et al. Epidemiology of catheter-related infections in adult patients receiving home parenteral nutrition: a systematic review. Clin Nutr 2013;32:16-26. 9. Winkler MF, DiMaria-Ghalili RA, Guenter P, Resnick HE, Robinson L, Lyman B, et al. Characteristics of a cohort of home parenteral nutrition patients at the time of enrollment in the Sustain registry. JPEN J Parenter Enteral Nutr 2015 May 13. [Epub ahead of print]. 10. Buchman AL, Opilla M, Kwasny M, Diamantidis TG, Okamoto R. Risk factors for the development of catheter-related bloodstream infections in patients receiving home parenteral nutrition. JPEN J Parenter Enteral Nutr 2014;38:744-9. 11. Pieroni KP, Nespor C, Ng M. Evaluation of ethanol lock therapy in pediatric patients on long-term parenteral nutrition. Nutr Clin Pract 2013;28:226-31. 12. Nguyen GC, Munsell M, Brant SR, LaVeist TA. Racial and geographic disparities in the use of parenteral nutrition among inflammatory bowel disease inpatients diagnosed with malnutrition in the United States. JPEN J Parenter Enteral Nutr 2009;33:563-8. 13. Mohammed A, Grant FK, Zhao VM, Shane AL, Ziegler TR, Cole CR. Characterization of posthospital bloodstream infections in children requiring home parenteral nutrition. JPEN J Parenter Enteral Nutr 2011;35:581-7. 14. Hortencio TD, Arendt BM, Teterina A, Jeejeebhoy KN, Gramlich LM, Whittaker JS, et al. Changes in home parenteral nutrition practice based on the Canadian home parenteral nutrition patient registry. JPEN J Parenter Enteral Nutr 2015 Sep 25. [Epub ahead of print]. 15. Botella-Carretero JI, Carrero C, Guerra E, Valbuena B, Arrieta F, Calañas A, et al. Role of peripherally inserted central catheters in home parenteral nutrition: a 5-year prospective study. JPEN J Parenter Enteral Nutr 2013;37:544-9. 16. Chopra V, O’Horo JC, Rogers MA, Maki DG, Safdar N. The risk of bloodstream infection associated with peripherally inserted central catheters compared with central venous catheters in adults: a systematic review and meta-analysis. Infect Control Hosp Epidemiol 2013;34:908-18. 17. O’Grady NP, Alexander M, Burns LA, Dellinger EP, Garland J, Heard SO, et al. Guidelines for the prevention of intravascular catheter-related infections. Am J Infect Control 2011;39(Suppl 1):S1-34. 18. Staun M, Pironi L, Bozzetti F, Baxter J, Forbes A, Joly F, et al. ESPEN Guidelines on parenteral nutrition: home parenteral nutrition (HPN) in adult patients. Clin Nutr 2009;28:467-79. 19. Cotogni P, Pittiruti M, Barbero C, Monge T, Palmo A, Boggio Bertinet D. Cather-related complications in cancer patients on home parenteral nutrition: a prospective study of over 51,000 catheter days. JPEN J Parenter Enteral Nutr 2013;37:375-83. 20. Bech LF, Drustrup L, Nygaard L, Skallerup A, Christensen LD, Vinter-Jense L, et al. Environmental risk factors for developing catheter-related bloodstream infections in home parenteral nutrition patients: a 6-year follow up study. JPEN J Parenter Enteral Nutr 2016;40:989-94. 21. Drews BB, Sanghavi R, Siegel JD, Metcalf P, Mittal NK. Characteristics of catheterrelated bloodstream infections in children with intestinal failure: implications for clinical management. Gastroenterol Nurs 2009;32:385-90. 22. Mermel LA, Alang N. Adverse effects associated with ethanol catheter lock solutions: a systematic review. J Antimicrob Chemother 2014;69:2611-9. 23. Kaiser Family Foundation. Health Reform Implementation Timeline, 2015. Available from: http://kff.org/interactive/implementation-timeline/. Accessed April 4, 2016.
Contents lists available at ScienceDirect
American Journal of Infection Control
American Journal of Infection Control
j o u r n a l h o m e p a g e : w w w. a j i c j o u r n a l . o r g
Major Article
Central venous catheter infections in home parenteral nutrition patients: Outcomes from Sustain: American Society for Parenteral and Enteral Nutrition’s National Patient Registry for Nutrition Care Vicki M. Ross PhD, RN, CNSC a, Peggi Guenter PhD, RN, FAAN a,*, Mandy L. Corrigan MPH, RD, CNSC, FAND a, Debra Kovacevich MPH, RN b, Marion F. Winkler PhD, RD, LDN, CNSC, FASPEN c, Helaine E. Resnick PhD, MPH d, Tina L. Norris PhD d, Lawrence Robinson PharmD, MS e, Ezra Steiger MD, FACS, FASPEN, AGAF f a
American Society for Parenteral and Enteral Nutrition, Silver Spring, MD University of Michigan Home Care Services, University of Michigan, Ann Arbor, MI c Rhode Island Hospital and Alpert Medical School of Brown University, Providence, RI d Silver Spring, MD e Collierville, TN f Digestive Disease Institute, Cleveland Clinic, Cleveland, OH b
Key Words: Database CLABSI
Background: Home parenteral nutrition (HPN) is a high-cost, complex nutrition support therapy that requires the use of central venous catheters. Central line-associated bloodstream infections (CLABSIs) are among the most serious risks of this therapy. Sustain: American Society for Parenteral and Enteral Nutrition’s National Patient Registry for Nutrition Care (Sustain registry) provides the most current and comprehensive data for studying CLABSI among a national cohort of HPN patients in the United States. This is the first Sustain registry report detailing longitudinal data on CLABSI among HPN patients. Objective: To describe CLABSI rates for HPN patients followed in the Sustain registry from 2011-2014. Methods: Descriptive, χ2, and t tests were used to analyze data from the Sustain registry. Results: Of the 1,046 HPN patients from 29 sites across the United States, 112 (10.7%) experienced 194 CLABSI events during 223,493 days of HPN exposure, for an overall CLABSI rate of 0.87 episodes/1,000 parenteral nutrition-days. Although the majority of patients were female (59%), adult (87%), white (75%), and with private insurance or Medicare (69%), CLABSI episodes per 1,000 parenteral nutrition-days were higher for men (0.69 vs 0.38), children (1.17 vs 0.35), blacks (0.91 vs 0.41), and Medicaid recipients (1.0 vs 0.38 or 0.39). Patients with implanted ports or double-lumen catheters also had more CLABSIs than those with peripherally inserted or central catheters or single-lumen catheters. Staphylococci were the most commonly reported pathogens. These data support findings of smaller studies about CLABSI risk for children and by catheter type and identify new potential risk factors, including gender, race, and insurance type. Conclusions: Additional studies are needed to determine effective interventions that will reduce HPN-associated CLABSI. © 2016 Association for Professionals in Infection Control and Epidemiology, Inc. Published by Elsevier Inc. All rights reserved.
* Address correspondence to Peggi Guenter, PhD, RN, FAAN, American Society for Parenteral and Enteral Nutrition, 8630 Fenton St. Suite 412 Silver Spring, MD 20910. E-mail address: [email protected] (P. Guenter). Sustain was supported, in part, by unrestricted scientific grants from Baxter Healthcare Corporation, the C. R. Bard Foundation, and the A.S.P.E.N. Rhoads Research Foundation. Conflicts of interest: None to report.
Assessments of high-risk, complex patient populations require large datasets or registries to generate detailed outcome data. Registries are especially useful to describe a real-world picture of patient populations to better understand their disease course.1 Recognizing the need for data that accurately describe the characteristics and outcomes of high-risk, complex patients receiving parenteral nutrition (PN) at home, the American Society for Parenteral and Enteral Nutrition (A.S.P.E.N.) launched Sustain: A.S.P.E.N.’s National Patient
0196-6553/© 2016 Association for Professionals in Infection Control and Epidemiology, Inc. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ajic.2016.06.028
V.M. Ross et al. / American Journal of Infection Control 44 (2016) 1462-8
Registry for Nutrition Care (Sustain registry) in 2011. The Sustain registry was designed to collect and compile data on adult and pediatric home parenteral nutrition (HPN) patients.2 The Sustain registry generates longitudinal observational data that can be used to describe and understand the HPN population in the United States and provide evidence for future research. Developed initially to provide intravenous nutrition for acutely ill hospitalized patients with intestinal failure, PN therapy became an option for home care in 1969. HPN allowed patients to receive this complex nutrition therapy in a familiar, comfortable environment. Along with the convenience of administering PN at home came the significant responsibility for patients and their family caregivers to provide direct care.3 Occurring both in the hospital and at home, central lineassociated bloodstream infections (CLABSIs) are among the most common and serious complications of PN therapy. For hospitalized patients there are standardized definitions and data collection methods to define and determine CLABSI that in turn allow for benchmarking infection rates, research activities, and reviews by the public and third-party payers. According to a 2013 National Healthcare and Safety Network Report, 4 CLABSI rates for hospitalized patients in nonburn units range from 0-1.2 per 1,000 catheterdays. Remarkably, there are no corresponding standardized definitions or mandatory data reporting for HPN patients. The number of HPN patients in the United States is currently unknown. A 2010 survey from the National Home Infusion Association estimated that 33,000 people in the United States require HPN. The same survey estimated that annual spending for HPN was $652 million.5 These estimates are likely lower than the actual numbers of HPN patients because not all HPN providers participated in the National Home Infusion Association survey. In the United States, there are few large datasets describing characteristics and clinical outcomes of HPN patients. Historical data from 19851992 are available from the Oley-A.S.P.E.N. Information System and the North American Home Parenteral and Enteral Patient Registry has some data from 1992-1994.6 In 2005, a home care company published outcomes from a 5-year retrospective evaluation of their experience with 4,540 HPN patients.7 These databases are no longer current or available for public use, leaving significant gaps in knowledge concerning outcomes of HPN patients in the United States. HPN rates of infections reported in a systematic review of published studies ranged from 0.38-4.58 episodes per 1,000 catheterdays. The lack of standardized definitions, data collection methods, and reporting mechanisms for HPN patients contribute to the wide range in reported rates.8 There is an obvious need for current data on CLABSI rates in HPN patients. The Sustain registry database provides the most comprehensive and current HPN patient information in the United States. The purpose of this longitudinal study was to describe the CLABSI rate and characteristics for HPN patients followed in the Sustain registry from 2011-2014. METHODS Nutrition support professionals who provided HPN services in outpatient or home-care sites were invited to participate in the Sustain registry by providing de-identified observational or chart data about their HPN patients. To ensure protection of human subjects, A.S.P.E.N. received approval for the overall conduct of the Sustain registry from the New England Institutional Review Board and participant providers at individual sites obtained requisite approvals from their local institutional review boards. As required by their local institutional review boards, participating providers obtained written consent or waivers of consent from their HPN patients. Additional details about the development and conduct of the Sustain registry are provided in earlier publications.2,9 The registry system has a built-in quality control
1463
program to alert site participants and registry administration if duplicate patients are entered. In the few instances that this occurred, the registry administrator worked out this issue with the sites. Baseline observations from chart data collection included demographic, HPN, and catheter characteristics. Follow-up observations from chart data included information about occurrence of CLABSIs and pathogens associated with these events. When CLABSIs were reported, participating sites were directed to provide information about the pathogen associated with the infection. HPN duration was defined as the time in days HPN patients were followed in the Sustain registry from baseline to the final followup visit or patient contact. Because Sustain is an observation type registry, the time frames between the baseline and 1 or more followup visits were determined by participating providers based on their clinical process. In most cases, a follow-up encounter corresponded with an episode of care (eg, outpatient visit, hospitalization, and routine visit), but some encounters involved telephone contact. This study reports CLABSI data for patients who were enrolled in the Sustain registry between August 2011 and February 2014 and who had at least 1 follow-up visit or contact through August 2014. Definitions Patients starting HPN therapy within 90 days of baseline data collection were defined as “new,” and their PN-days were calculated from the actual HPN start date. Patients receiving HPN therapy for >90 days before baseline data collection were defined as “existing,” and their PN-days were calculated from the date of baseline enrollment in the registry. No attempt was made to ascertain the exact date of HPN initiation for the “existing” patients. Thus, PNdays and CLABSI rates reflect the time patients were receiving HPN and followed in the Sustain registry. CLABSIs were defined as primary bloodstream infections in patients who had a central venous catheter (CVC) within the 48hour period before the development of the bloodstream infection and not related to infections at other body sites. This definition was shared with each site during Sustain registry training and placed on the help pages in the online data collection forms. Each participating site determined whether a patient’s infection met the criteria for CLABSI. The overall CLABSI rate was calculated as the number of CLABSI events per 1,000 PN-days. Patient-level characteristics were also assessed in terms of presence or absence of at least 1 CLABSI, and these rates were calculated on the patient level. Data analysis Data analyses included the 1,046 patients who were enrolled in the Sustain registry between August 2011 and February 2014, and who had at least 1 follow-up visit or contact through August 2014. Overall CLABSI rates were calculated as events per unit time and corresponding rates for demographic and clinical characteristics were calculated on the patient level among those with 1 or more CLABSI episodes. To determine whether CLABSI status differed by demographic, clinical, or catheter-related characteristics, χ2 tests for categorical and t tests for continuous data were used. Analyses were conducted using SAS (SAS Institute Inc, Cary, NC). RESULTS Site characteristics Twenty-one of 29 participating sites were hospitals and the remaining sites were home infusion companies. Of the hospital-based sites, 35% were academic medical centers with adult patients alone or adult and pediatric patients, 60% were free-standing academic
1464
V.M. Ross et al. / American Journal of Infection Control 44 (2016) 1462-8
children’s hospitals, and 5% were community hospitals. Hospitals were located in 26 cities across 15 states. The size of hospital-based sites varied from 430-1,000 beds in academic medical centers, and 85427 beds in children’s hospitals. Sixty-nine percent of HPN patients, of whom the majority were adults, were enrolled by medical centers and hospitals, 19% were from home care or infusion service providers with both adult and pediatric patients, and the remaining 11% were enrolled at children’s hospitals. Half of the sites enrolled all “new” HPN patients and 60% approached all “existing” HPN patients to participate. Most sites (83.3%) consecutively enrolled and followed HPN patients during their participation in the Sustain registry. Sample characteristics Of all 1,251 Sustain registry patients with the baseline data previously described,9 1,046 had at least 1 follow-up visit or contact with their participating site provider and met eligibility criteria for this longitudinal, descriptive study. In this observational study, although not common it was possible for patients to be entered into the registry and return 1 day later for their last recorded follow-up visit. Mean HPN duration was 214 days (range, 1-946 days). Adult patients received HPN an average of 202 days, and pediatric patients received HPN an average of 291 days. There were 194 CLABSI events in 112 patients. The overall CLABSI rate was 0.87 episodes per 1,000 PN-days. Demographic characteristics Table 1 shows the demographic characteristics of the sample by CLABSI status. A common profile of a Sustain registry patient was an adult, white, woman. However, when comparing those who had at least 1 CLABSI to those with no infections for different demographic categories, children, nonwhites, and males accounted for 13.2% of the overall sample, but 42% of those experienced 1 or more CLABSI and only 9.7% had no CLABSI. Similarly, blacks accounted for 12% of the sample but 20.5% of those who had at least 1 CLABSI. These differences were also reflected in the higher infection rates for children, blacks, and males.
Among patients with data about insurance coverage, most had private or commercial insurance, or Medicare, followed by Medicaid alone or in combination. Insurance coverage differed by infection status. Nearly 47% of patients with at least 1 infection reported Medicaid coverage. Patients receiving Medicaid had CLABSI rates that were more than twice the rate of those with Medicare or private pay (1.0 vs 0.38 or 0.39/1,000 HPN-days). Clinical characteristics The 5 most common diagnoses reflected some degree of bowel dysfunction, with CLABSI occurring significantly more frequently in the short bowel group (Table 2). Unexpectedly, the 18.7% reporting gastrointestinal cancer diagnosis accounted for <5% of patients with CLABSI. Data about the presence or absence of small and large bowel stomas were collected for 299 patients with short bowel syndrome. Of these patients, 57.3% had a small bowel stoma and 8.4% had a large bowel stoma. There was no statistically significant difference in the number of infections between patients with or without stomas. More than half (60.8%) of patients in the sample had HPN duration of 6 months or less and 21% were followed for a year or more while receiving HPN. CLABSIs were significantly more frequent in the patients categorized as longer HPN duration; yet, when accounting for the number of days of exposure, the CLABSI rates per 1,000 PN-days were lowest for those followed a year or longer. More than 90% of patients in this study received daily infusions of HPN and 35.2% received intravenous lipid emulsions daily. There were no statistically significant differences for the occurrence of CLABSI when comparing patients receiving daily HPN to those receiving HPN fewer times per week or for patients receiving daily lipids to those receiving lipids fewer times per week. Catheter characteristics Ninety percent of patients administered their HPN through a tunneled catheter (47%) or a peripherally inserted central catheter (PICC)
Table 1 Selected demographic characteristics by infection status at follow-up, Sustain: American Society for Parenteral and Enteral Nutrition’s National Patient Registry for Nutrition Care
Variable Total Gender Women Men Age group (y) < 18 ≥ 18 Race White Black Other Ethnicity Hispanic Non-Hispanic Insurance Medicare Medicaid Private Personal pay Medicare supplement Other
Entire sample n (%) 1,046 (100) 1,045 (100) 614 (58.8) 431 (41.2) 1,046 (100) 138 (13.2) 908 (86.8) 1,046 (100) 780 (74.6) 125 (12) 141 (13.5) 927 (100) 63 (6.0) 864 (93.2) 982 (100) 224 (22.8) 196 (20.0) 500 (50.9) 15 (1.5) 22 (2.2) 25 (2.5)
Total parenteral nutrition-days
CLABSI rate* ‡
223,493
0.87
136,112 87,200
0.38 0.69
40,133 183,360
1.17 0.35
171,528 25,217 26,748 175,505 13,270 162,235
0.41 0.91 0.71 0.90 0.51
44,669 48,016 80,305 1,634 5,280 3,107
0.38 1.00 0.39 0.61 0.19 1.61
Patients with 1 or more CLABSI n (%)
Patients with no CLABSI n (%)
112 (10.7) 112 (100) 52 (46.2) 60 (53.6) 112 (100) 47 (42) 65 (58) 112 (100) 70 (62.5) 23 (20.5) 19 (7.0) 95 (100) 12 (12.6) 83 (87.4) 103 (100) 17 (16.5) 48 (46.6) 31 (30.1) 1 (1.0) 1 (1.0) 5 (4.8)
934 (89.3) 933 (100) 562 (60.2) 371 (39.8) 934 (100) 91 (9.7) 843 (90.3) 934 (100) 710 (76.0) 102 (10.9) 122 (13.1) 832 (100) 51 (5.5) 781 (93.9) 879 (100) 207 (23.5) 148 (16.8) 469 (53.4) 14 (1.6) 21 (2.4) 20 (2.3)
CLABSI, central line-associated bloodstream infection. *CLABSI rates are presented at the patient level unless otherwise indicated. †P values reflect χ2 or Fisher exact test, as appropriate. ‡ Overall CLABSI rate for the sample is based on total number of CLABSI events divided by total parenteral nutrition-days.
P value† .0050
< .0001
.0035
.0171
.0886 < .0001 < .0001 1.0000 .5004 .1766
V.M. Ross et al. / American Journal of Infection Control 44 (2016) 1462-8
1465
Table 2 Selected clinical characteristics by infection status at follow-up, Sustain: American Society for Parenteral and Enteral Nutrition’s National Patient Registry for Nutrition Care Entire sample n (%) Total Top-5 diagnoses Short bowel syndrome Gastrointestinal cancer Crohn’s disease Gastromotility/pseudo-obstruction disease Gastrointestinal bypass for obesity Duration of parenteral nutrition at follow-up (d) ≤ 90 91-180 181-365 > 365
Total parenteral nutrition-days
CLABSI rate*
Patients with 1 or more CLABSI n (%)
Patients with no CLABSI n (%)
1,046 (100)
223,493
0.87‡
112 (10.7)
934 (89.3)
299 (28.6) 196 (18.7) 171 (16.3) 116 (11.1) 75 (7.2)
108,378 25,823 48,727 35,077 10,120
0.48 0.19 0.25 0.48 0.1
52 (46.2) 5 (4.5) 12 (10.7) 17 (15.2) 1 (0.9)
247 (26.4) 191 (20.4) 159 (17) 99 (10.6) 74 (7.9)
485 (46.4%) 172 (16.4%) 166 (15.9%) 223 (21.3%)
20,916 22,328 42,901 137,348
0.67 0.67 0.68 0.39
14 (12.5%) 15 (13.4%) 29 (25.9%) 54 (48.2%)
P value†
< .0001 < .0001 .088 .1448 .0064 < .0001
471 (50.4%) 157 (16.8%) 137 (14.7%) 169 (18.1%)
CLABSI, central line-associated bloodstream infection. *CLABSI rates are presented on the patient level unless otherwise indicated. †P values reflect χ2 or Fisher exact test, as appropriate. ‡ Overall CLABSI rate for the sample is based on total number of CLABSI events divided by total parenteral nutrition-days.
Table 3 Catheter type, related care, and pathogens by infection status at follow-up, Sustain: American Society for Parenteral and Enteral Nutrition’s National Patient Registry for Nutrition Care
Variable Total Catheter type Peripherally inserted central catheter Port Tunneled Other Number of lumens Single Double Triple Use of ethanol lock No Yes Use of anti-infectives No Yes Dressing changed by: Patient and caregiver Caregiver Patient Visiting nurse
Entire sample n (%)
Total parenteral nutrition-days
CLABSI rate*
Patients with 1 or more CLABSI n (%)
Patients with no CLABSI n (%)
1,046 (100) N = 1,045 449 (42.9) 93 (8.9) 492 (47) 11 (1.1)
223,493
0.87‡
112 (10.7)
934 (89.3)
53,978 18,131 150,087 1,297
0.41 0.66 0.51 0.77
22 (19.6) 12 (10.7) 77 (68.8) 1 (0.9)
427 (45.7) 81 (8.7) 415 (44.4) 10 (1.1)
481 (46) 538 (51.4) 27 (2.6)
73,749 147,963 1,781
0.38 0.57 0.00
28 (25.0) 84 (75.0) 0 (0.0)
453 (48.5) 454 (48.6) 27 (2.9)
762 (72.8) 284 (27.2)
142,298 81,195
0.48 0.54
68 (60.7) 44 (39.3)
694 (74.3) 240 (25.7)
788 (75.3) 258 (24.7) 568 (100) 15 (2.6) 102 (18.0) 63 (11.1) 388 (68.3)
174,423 49,070
0.50 0.51
1,978 34,631 32,844 60,018
0.00 0.75 0.27 0.50
87 (77.7) 25 (22.3) 65 (11.4) 0 (0.0) 26 (40) 9 (13.8) 30 (46.2)
701 (75.1) 233 (24.9) 503 (88.6) 15 (3.0) 76 (15.1) 54 (10.7) 358 (72.2)
P value† < .0001
< .0001
.0022
.5425
< .0001†
CLABSI, central line-associated bloodstream infection. *CLABSI rates are presented on the patient level unless otherwise indicated. †P values reflect χ2 or Fisher exact test, as appropriate. ‡Overall CLABSI rate for the sample is based on total number of CLABSI events divided by total parenteral nutrition-days.
(42.9%). The remaining 9% reported using subcutaneous infusion ports. Table 3 contains data about catheter types and related care. Among patients with at least 1 CLABSI, 19.6% used a PICC compared with 68.8% of those with a tunneled catheter. Ninety-seven percent of patients had either a single- or doublelumen CVC. The CLABSI events were statistically more for the doublelumen CVC compared with the single-lumen CVC. There were no CLABSIs reported for the 27 patients with triple-lumen catheters. The use of ethanol locks were reported in 27.2% of patients who accounted for 39.3% of the infections, whereas the remaining 72.8% of the group did not use ethanol locks and accounted for 60.7% of the infections. Data from the medication inventory revealed that 24.7% of patients received anti-infective agents (oral or intravenous, antibiotics, or antifungal agents) at baseline. There was no statistically significant difference (P = .54) for CLABSI events between patients who did and did not report use of anti-infective agents. Information about who was changing CVC dressings was available for 568 patients. Data were not reported or missing for the
remaining patients. CLABSI rates were highest among patients for whom caregivers changed the dressings alone (0.75 events per 1,000 PN-days). There were no CLABSIs reported for the 15 HPN patients who reported dressings changed by patients and a caregiver. Identified pathogens There were 194 CLABSI episodes in 112 patients, and 96 patients had the pathogens identified. Reported pathogens included gram-positive, gram-negative, and fungal types. Sixteen percent of pathogens were listed as “other” without further detail. Figure 1 provides additional detail about the specific pathogens reported. DISCUSSION AND LIMITATIONS Nearly 11% of patients in our study experienced one or more CLABSI, supporting the concern that CLABSIs continue to remain for HPN patients. Furthermore, the most vulnerable and potentially
1466
V.M. Ross et al. / American Journal of Infection Control 44 (2016) 1462-8
Fig 1. Central line-associated bloodstream infection pathogens.
disenfranchised patients—children, minorities, and those receiving public assistance—experienced significantly higher numbers of infections. Our findings support the results of smaller or singlesite studies and provide additional evidence about the need to improve HPN care. Smaller studies with a variety of defining criteria for HPNrelated infections contribute to the wide range of reported infections in the literature.8 Although the exact number of HPN patients in the United States is not known, the overall CLABSI at the patient level, the rate for patients in the Sustain registry (0.87 per 1,000 PN-days; pediatric rate is 1.17 per 1,000 catheter-days and the adult rate is 0.35 per 1,000 PN-days) is within the range of other HPN noncancer pediatric studies of 0.19 and 2.14 per 1,000 catheter-days.8 Similarly, our CLABSI rate of 1.17 per 1,000 HPN-days is within the range reported by others (0.8010 and 9.811 infections per 1,000 catheter-days). Black and Hispanic patients in the Sustain registry experienced more than twice the rate of infection compared with whites. This is among of the first studies to highlight the potential risk, either directly or indirectly, that race has on HPN risk. Nguyen et al12 in a study of inpatients with inflammatory bowel disease, a condition associated with PN use, found malnourished African Americans were less likely than whites to receive PN (odds ratio, 0.67; 95% confidence interval, 0.50-0.89) and Hispanics had higher rates of catheter-related complications (5.1% vs 12.2%; P = .04). Whether race is directly or indirectly related, this new information underscores the importance of determining why these differences occur to be able to develop effective interventions and culturally appropriate educational materials. In our study, Medicaid patients had more than twice the rate of infection as those with private insurance. In at least 1 other study, the CLABSI rates were higher for pediatric patients with Medicaid insurance.13 More research is needed to determine and address the specific factors related to these findings. It is possible that Medicaid is a proxy for socioeconomic status. Alternately, it may reflect patient age, with higher infection rates among children who also use Medicaid. Notably, the CLABSI rates for Medicare and private insurance (0.38 vs 0.39/1,000 HPN-days) in our study were almost identical, suggesting chronological age of adults may have minimal influence on infection risk. The specific differences in CLABSI rates
by insurance coverage cannot be determined by registry data. These findings support the need for further study. HPN is considered a long-term therapy requiring a long-term catheter. Almost half of patients in the Sustain registry were using a PICC, with a higher number of CLABSIs in the implanted or and tunneled catheter groups. The proportion of PICCs reported in our study align with a study of trends using the Canadian HPN patient registry, where there was an increased use of PICCs from the initial 21.6% during the 2005-2008 period to 54.9% during the 20112014 period.14 There are at least 2 HPN studies suggesting PICCs are not associated with a higher risk of infection when compared with other CVCs. In a 5-year prospective study of HPN patients there were no infections reported for the PICC group and 44 and 20 per 1,000 catheter-days for the implanted and tunneled catheter groups, respectively.15 In another prospective study of HPN cancer patients, there were no infections reported for the 65 patients with a PICC and infection rates of 0.97 and 0.45 per HPN-days for the tunneled and implanted catheter groups, respectively.15 In a metaanalysis of 23 studies and more than 57,000 patients requiring CVCs for varied purposes, PICCs were less likely to be associated with CLABSI than other CVCs.16 Patients in the Sustain registry with double-lumen catheters experienced more CLABSIs than those with single-lumen catheters. These findings support recommendations to use a single-lumen CVC or the fewest number of lumens necessary to manage patient care.17,18 Surprisingly, there were no CLABSIs reported for the 27 Sustain registry patients using triple-lumen catheters. The patients in this group had a shorter average follow-up time compared with the doublelumen group (66 days vs 275 days). Further research is needed to determine other factors in addition to the number of catheter lumens that increase risk for CLABSI. For the 194 CLABSI episodes in 112 patients, 96 patients had pathogens identified. The majority of pathogens could be categorized as gram-positive, gram-negative, or fungal species. However, of the patients with identified pathogens, 16% were listed as “other” with no more specific identification. Our findings are similar to the pattern described by others who found the majority of infections were gram-positive infections followed by gram-negative and fungal infections.8,13 Registry data did not contain detailed information about
V.M. Ross et al. / American Journal of Infection Control 44 (2016) 1462-8
the antibiotic susceptibility of the infecting organisms; however, the general pattern of pathogens reinforces the need for research that focuses on the prevention or eradication of these specific pathogens. Furthermore, given the number of patients in our study reporting anti-infective therapy at baseline, more detailed studies about antibiotic resistance and potentially changing patterns of pathogens are needed. The primary diagnoses of Sustain registry patients reflected some degree of intestinal impairment. Surprising findings from this study were the reported CLABSI events for patients with cancer, stomas, and ethanol lock therapy. The low CLABSI rate for cancer patients in the Sustain registry was somewhat surprising given the higher infection rates of 0.54-6.8 per 1,000 catheter-days reported for HPN cancer patients in other studies.8 Only 5 of 196 patients in the Sustain registry with gastrointestinal cancer as the primary diagnosis experienced a CLABSI. It was outside the purpose of the Sustain registry to determine the exact type of cancer or cancer treatment. Our results were similar to the results of a prospective study of 254 European HPN cancer patients, the majority with solid tumor malignancies, reporting an overall CLABSI rate of 0.35 per 1,000 catheter-days.19 Because of the close proximity of catheters exiting the body and the enteric flora found at stoma sites, there is concern for crosscontamination and increased risk for infections when HPN patients have abdominal ostomies. In a 6-year observational cohort of 136 adult HPN patients, those with stomas and tunneled catheters experienced a decreased time to first CLABSI.20 Drews et al21 also found ostomies to be a risk factor for pediatric patients. Counter to the findings of others and similar to findings in the Sustain registry, Buchman et al10 found similar catheter-related infection rates for HPN patients with and without ostomies (0.34 vs 0.4 infections per 1,000-days). A more detailed study of patients with ostomies is needed to determine whether there are coexisting characteristics for some patients with ostomies that increase or ameliorate their risk of infection. For patients with recurrent infections, the Centers for Disease Control and Prevention recommends use of prophylactic antimicrobial lock solutions, of which ethanol is an example.17 More than a quarter of patients upon entry into the Sustain registry reported using ethanol locks, suggesting at least those using the lock solution may have experienced 1 or more CLABSI before enrollment. CLABSI events were significantly higher for patients with ethanol locks while followed in the Sustain registry compared with those who did not report ethanol lock at baseline. There is a paucity of data from prospective, randomized clinical trials about the most effective use of ethanol lock (eg, ideal catheter dwell time or frequency, strength, and dose).17,22 A detailed description about the use of ethanol locks was not a part of the Sustain registry and more research is needed to understand the use and effects of ethanol locks. Compared against their previous HPN experience, the ethanol group may have had fewer CLABSIs but continued to experience some CLABSIs while in the registry. Additionally, there was a national shortage of ethanol that occurred during the Sustain registry data collection and patients who were initially receiving the therapy may have been forced to reduce or eliminate the dose or frequency of the ethanol. Findings of this study are limited by the recruitment of volunteer sites and the use of data typically recorded by participating providers. The actual number of HPN patients is not known and it is difficult to determine how closely the volunteer sites approximated the actual HPN population or the bias that might have been introduced. Multiple attempts were made to communicate with and invite all HPN providers to participate in the Sustain registry. Although the Sustain registry did not have 100% participation, participants were located throughout the major regions of the United States and from both large and small sites, providing a more robust sample than what could otherwise be obtained from a single site or provider.
1467
Alternate explanations for some of our findings may be related to the noncompulsory recruitment or the detail of data routinely collected by participant providers and may not reflect the total HPN population. The purpose of the Sustain registry is to collect longitudinal, observational data about HPN patients that can be used in the development of future intervention testing with the ultimate goal of improving HPN care. Except to provide benchmarking data to the individual sites as a benefit of participating, there was no intent to evaluate the specific care provided by individual sites. The design of the Sustain registry does not allow for randomization or the testing of specific interventions. There were multiple checks within the Sustain registry system to avoid the same patient data from being entered twice by different providers and to address out-of-range or missing data. These checks are described in earlier publications.2,9 Sustain registry data are limited to information routinely collected by providers. Few data were missing for demographic variables such as age or gender, for example, but for data not routinely observed such as the person or persons changing dressings, a large portion of the data were missing. For some data that did not fall neatly into discrete categories such as multiple insurance types or anti-infective agents, recategorizing into nominal levels of yes or no most likely introduced some error in the analyses. Modeling data using a number of predictor variables as a function of HPN-days was discussed and determined not to be appropriate for this study for a couple of reasons. First, the purpose of our study was to provide descriptive data for future intervention testing research, so we opted to provide unadjusted rates without making statistical inferences. Second, the heterogeneity of the HPN population coupled with missing data for information that was not routinely collected by providers would result in an increasing number of cases lost as the number of predictor variables was added. For example, the number of CLABSI events for the pediatric population was significantly different from the adult population, and the unadjusted CLABSI rates were 3 times those of the adult population. To adjust rates based on modeling of predictor variables would have required a theoretical framework for selecting the appropriate predictor variables and a sample size that was large enough with complete data to provide adequate power. This study reflects HPN patients enrolled in the Sustain registry from 2011-2014. Trends in patient characteristics and conditions may change over time, and those changes may influence the data and study findings. For example, the number of PICCs used by HPN patients enrolled in the Sustain registry were surprising, but may reflect a trend found in the Canadian HPN registry.14 There is no universal health insurance plan in the United States and insurance plans even within the same carrier may differ, resulting in variable experiences for HPN patients. Implementation of the Affordable Care Act, which requires all US citizens to have insurance, began January 1, 2014.23 Data collection for this study began almost 3 years before and was almost complete before implementation of the Affordable Care Act provision. The influence of this provision on the health of HPN patients has yet to be determined, but the dramatic difference in CLABSI rates between those with and those without private or Medicare insurance underscores the need for further study. Conclusions The Sustain registry provides the most current and comprehensive data for HPN patients in the United States. Generalizability of the results are strengthened by standardized data collection methods and a large sample from providers across the United States. The findings from this study indicate HPN-associated CLABSIs are a continuing threat, especially to the most vulnerable populations (ie, young, minorities, and the poor). As in acute care, there is need for
1468
V.M. Ross et al. / American Journal of Infection Control 44 (2016) 1462-8
evidence-based interventions that will drive the CLABSI rate for HPN patients to zero. Unlike acute-care settings with hospital-based protocols used by trained health care personnel, safe and effective HPN therapy relies on patients and their family caregivers who must be able to receive, understand, and implement evidence-based information about HPN care. It is imperative that nutrition support teams, in particular physicians and nurses, convey and ensure implementation of best practices for insertion and care of CVCs when working with HPN patients and their families. Future research should examine the effects of sterile catheter insertion techniques, providing relevant educational information to HPN patients and their family caregivers about aseptic care. Continually providing feedback and support can provide the framework for lowering infection rates. Acknowledgment The authors thank the Sustain registry coordinator Katy Hanley and the data collection contributions of the following clinical sites: Arnold Palmer Hospital for Children (Orlando, FL), CarePoint Partners (Dallas, TX; Providence, RI; and Tampa, FL), Boston Children’s Hospital (Boston, MA), Children’s Hospital of Michigan (Detroit, MI), Children’s Mercy Hospital (Kansas City, MO), Children’s Hospital of Orange County (Orange, CA), Cleveland Clinic (Cleveland, OH), University of Pennsylvania Health System Clinical Nutrition Support Service (Philadelphia, PA), Connecticut Children’s Medical Center (Hartford, CT), Cook Children’s Medical Center (Fort Worth, TX), Equinox Healthcare Inc (Ellicott, MD), Geisinger Medical Center (Danville, PA), Indiana University Health Home Care (Indianapolis, IN), Intermountain Medical Center (Murray, UT), Le Bonheur Children’s Hospital (Memphis, TN), Nation’s Home Infusion LLC (Owings Mills, MD), Nationwide Children’s Hospital (Columbus, OH), Pediatric Home Service (Roseville, MN), Rady Children’s Hospital (San Diego, CA), Rhode Island Hospital and Hasbro Children’s Hospital (Providence, RI), Riley Hospital for Children at Indiana University Health (Indianapolis, IN), Sharp Home Infusion and Central Pharmacy Services (San Diego, CA), Spectrum Health Butterworth (Grand Rapids, MI), Texas Children’s Hospital Baylor College of Medicine (Houston, TX), ThriveRx (Cincinnati, OH), University of Michigan (Ann Arbor, MI), and University of Virginia Health System (Charlottesville, VA). References 1. Gliklich R, Dreyer N, Leavy M, editors. Registries for evaluating patient outcomes: a user’s guide. 3rd ed. AHRQ Publication No. 13(14)-EHC111. Two volumes. Rockville (MD): Agency for Healthcare Research and Quality; 2010. Available from: http://www.effectivehealthcare.ahrq.gov/registries-guide-3.cfm. Prepared by the Outcome DEcIDE Center [Outcome Sciences, Inc., a Quintiles company] under Contract No. 290 2005 00351 TO7. 2. Guenter P, Robinson L, DiMaria-Ghalili RA, Lyman B, Steiger E, Winkler MF. Development of SustainTM: A.S.P.E.N.’s National Patient Registry for Nutrition Care. JPEN J Parenter Enteral Nutr 2012;36:399-406.
3. Ayers P, Guenter P, Holcombe B, Plogsted S. Home parenteral nutrition. In: A.S.P.E.N. parenteral nutrition handbook. 2nd ed. Silver Spring (MD): American Society for Parenteral and Enteral Nutrition; 2014:247-85 [chapter 10]. 4. Dudeck MA, Edwards JR, Allen-Bridson K, Gross C, Malpiedi PJ, Peterson KD, et al. National healthcare safety network report, data summary 2013, device associated module. Am J Infect Control 2015;43:206-21. 5. National Home Infusion Association (NHIA). Phase I: 2010 NHIA provider survey comprehensive aggregate analysis report. Alexandria (VA): National Home Infusion Association; 2011. 6. Howard L. Home parenteral nutrition: survival, cost, and quality of life. Gastroenterology 2006;130(Suppl 1):S52-9. 7. Ireton-Jones C, DeLegge M. Home parenteral nutrition registry: a five-year retrospective evaluation of outcomes of patients receiving home parenteral nutrition support. Nutrition 2005;21:156-60. 8. Dreesen M, Foulon V, Spriet I, Goossens GA, Hiele M, De Pourcq L, et al. Epidemiology of catheter-related infections in adult patients receiving home parenteral nutrition: a systematic review. Clin Nutr 2013;32:16-26. 9. Winkler MF, DiMaria-Ghalili RA, Guenter P, Resnick HE, Robinson L, Lyman B, et al. Characteristics of a cohort of home parenteral nutrition patients at the time of enrollment in the Sustain registry. JPEN J Parenter Enteral Nutr 2015 May 13. [Epub ahead of print]. 10. Buchman AL, Opilla M, Kwasny M, Diamantidis TG, Okamoto R. Risk factors for the development of catheter-related bloodstream infections in patients receiving home parenteral nutrition. JPEN J Parenter Enteral Nutr 2014;38:744-9. 11. Pieroni KP, Nespor C, Ng M. Evaluation of ethanol lock therapy in pediatric patients on long-term parenteral nutrition. Nutr Clin Pract 2013;28:226-31. 12. Nguyen GC, Munsell M, Brant SR, LaVeist TA. Racial and geographic disparities in the use of parenteral nutrition among inflammatory bowel disease inpatients diagnosed with malnutrition in the United States. JPEN J Parenter Enteral Nutr 2009;33:563-8. 13. Mohammed A, Grant FK, Zhao VM, Shane AL, Ziegler TR, Cole CR. Characterization of posthospital bloodstream infections in children requiring home parenteral nutrition. JPEN J Parenter Enteral Nutr 2011;35:581-7. 14. Hortencio TD, Arendt BM, Teterina A, Jeejeebhoy KN, Gramlich LM, Whittaker JS, et al. Changes in home parenteral nutrition practice based on the Canadian home parenteral nutrition patient registry. JPEN J Parenter Enteral Nutr 2015 Sep 25. [Epub ahead of print]. 15. Botella-Carretero JI, Carrero C, Guerra E, Valbuena B, Arrieta F, Calañas A, et al. Role of peripherally inserted central catheters in home parenteral nutrition: a 5-year prospective study. JPEN J Parenter Enteral Nutr 2013;37:544-9. 16. Chopra V, O’Horo JC, Rogers MA, Maki DG, Safdar N. The risk of bloodstream infection associated with peripherally inserted central catheters compared with central venous catheters in adults: a systematic review and meta-analysis. Infect Control Hosp Epidemiol 2013;34:908-18. 17. O’Grady NP, Alexander M, Burns LA, Dellinger EP, Garland J, Heard SO, et al. Guidelines for the prevention of intravascular catheter-related infections. Am J Infect Control 2011;39(Suppl 1):S1-34. 18. Staun M, Pironi L, Bozzetti F, Baxter J, Forbes A, Joly F, et al. ESPEN Guidelines on parenteral nutrition: home parenteral nutrition (HPN) in adult patients. Clin Nutr 2009;28:467-79. 19. Cotogni P, Pittiruti M, Barbero C, Monge T, Palmo A, Boggio Bertinet D. Cather-related complications in cancer patients on home parenteral nutrition: a prospective study of over 51,000 catheter days. JPEN J Parenter Enteral Nutr 2013;37:375-83. 20. Bech LF, Drustrup L, Nygaard L, Skallerup A, Christensen LD, Vinter-Jense L, et al. Environmental risk factors for developing catheter-related bloodstream infections in home parenteral nutrition patients: a 6-year follow up study. JPEN J Parenter Enteral Nutr 2016;40:989-94. 21. Drews BB, Sanghavi R, Siegel JD, Metcalf P, Mittal NK. Characteristics of catheterrelated bloodstream infections in children with intestinal failure: implications for clinical management. Gastroenterol Nurs 2009;32:385-90. 22. Mermel LA, Alang N. Adverse effects associated with ethanol catheter lock solutions: a systematic review. J Antimicrob Chemother 2014;69:2611-9. 23. Kaiser Family Foundation. Health Reform Implementation Timeline, 2015. Available from: http://kff.org/interactive/implementation-timeline/. Accessed April 4, 2016.