Abstract
Background and objectives
Infection-related hospitalizations have increased dramatically over the last 10 years in patients receiving in-center hemodialysis. Patient and dialysis facility characteristics associated with the rate of infection-related hospitalization were examined, with consideration of the region of care, rural-urban residence, and socioeconomic status.
Design, setting, participants, & measurements
The US Renal Data System linked to the American Community Survey and Rural-Urban Commuting Area codes was used to examine factors associated with hospitalization for infection among Medicare beneficiaries starting in-center hemodialysis between 2005 and 2008. A Poisson mixed effects model was used to examine the associations among patient and dialysis facility characteristics and the rate of infection-related hospitalization.
Results
Among 135,545 Medicare beneficiaries, 38,475 (28%) had at least one infection-related hospitalization. The overall rate of infection-related hospitalization was 40.2 per 100 person-years. Age ≥85 years old, cancer, chronic obstructive pulmonary disease, inability to ambulate or transfer, drug dependence, residence in a care facility, serum albumin <3.5 g/dl at dialysis initiation, and dialysis initiation with an access other than a fistula were associated with a ≥20% increase in the rate of infection-related hospitalization. Patients residing in isolated small rural compared with urban areas had lower rates of hospitalization for infection (rate ratio, 0.91; 95% confidence interval, 0.86 to 0.97), and rates of hospitalization for infection varied across the ESRD networks. Measures of socioeconomic status (at the zip code level), total facility staffing, and the composition of staff (percentage of nurses) were not associated with the rate of hospitalization for infection.
Conclusions
Patient and facility factors associated with higher rates of infection-related hospitalization were identified. The findings from this study can be used to identify patients at higher risk for infection and inform the design of infection prevention strategies.
Keywords: end-stage renal disease; hemodialysis; epidemiology and outcomes; hospitalization; humans; kidney failure, chronic; renal dialysis; risk factors; serum albumin; social class
Introduction
As of 2012, >400,000 adults were receiving hemodialysis for ESRD in the United States (1). Infection is a well recognized complication (2,3) and the second leading cause of hospitalization and death in adults receiving hemodialysis. Hospitalizations primarily for infection have increased 34% between 1993 and 2012 (1) and are associated with high rates of 30-day readmission and death (4). In recognition of these important clinical consequences, policy changes to the ESRD Quality Incentive Program (5) designed to lower infection rates were implemented, including reporting to the Centers for Disease Control and Prevention’s (CDC’s) National Healthcare Safety Network (NHSN) starting in payment year 2014 and the addition of NHSN bloodstream infections as a clinical measure starting in payment year 2016 (6). To develop the most effective and efficient infection prevention strategies, it is important to better understand patient– and facility–level risk factors for infection-related hospitalization. The primary objective of our study was to identify patient characteristics and dialysis facility characteristics associated with the risk of infection-related hospitalization among Medicare beneficiaries starting in-center hemodialysis, with consideration of the region of care, rural-urban residence, and socioeconomic status (SES) indicators. We examined both modifiable and nonmodifiable characteristics, because modifiable characteristics (e.g., vascular access type and staffing) could be actionable, and nonmodifiable factors may be useful for risk stratification.
Materials and Methods
Study Cohort and Data Collection
We used the US Renal Data System (USRDS) linked to the US Census Data from the American Community Survey (ACS) and Rural-Urban Commuting Area (RUCA) codes to retrospectively assemble a cohort of adults starting dialysis between January of 2005 and June of 2008. We limited the cohort to adults 18–100 years of age who survived the first 90 days of dialysis, had Medicare Parts A and B (and one or more institutional claims, which include dialysis-related claims, in the 180 days after study start), had an initial Medical Evidence Form (Centers for Medicare & Medicaid Services Form 2728) version 2005, and were receiving dialysis in a free–standing dialysis facility with known profit status on day 91 of dialysis. We have previously characterized this cohort in detail (7); for this study, we limited our examination to patients on in-center hemodialysis and further excluded patients with missing data on SES indicators (n=1941) or RUCA codes (n=454), who were receiving care in dialysis centers with missing dialysis facility volume (n=519), implausible facility data (i.e., reported data inconsistent with subsequent years of data) (n=9), or zero or missing nursing staff data reported the year of study entry (n=2105); or who were hospitalized throughout the entire duration of his/her follow-up (n=92) (Supplemental Figure 1).
We incorporated SES data collected from the 2007–2011 ACS 5-year estimates for the zip code–tabulated area (ZCTA) (8). We linked the ZCTA-level data to patient–level zip codes at the start of dialysis using the Dartmouth Atlas 2009 zip code to ZCTA cross–walk file (9). To assign residence to a rural or urban area, we used the zip code approximation of the RUCA codes using the zip code–level RUCA version 2.0 files available through the Washington, Wyoming, Alaska, Montana, Idaho Rural Health Research Center at the University of Washington (10).
We collected the following patient characteristics from the Medical Evidence Form: demographics, coexisting illnesses, inability to ambulate or transfer, residence in a care facility, tobacco use, alcohol dependence, drug dependence, serum albumin concentration, height and weight (used to calculate Quételet [body mass] index [BMI]), eGFR calculated from the abbreviated Modification of Diet in Renal Disease equation (11), vascular access type used at the first outpatient dialysis session, and nephrology care before dialysis initiation (classified as unknown, none, or <6, 6–12, or >12 months). We ascertained the ESRD network, and we classified patient rural-urban location as follows: urban, large rural town (micropolitan), small rural town, and isolated small rural town (12). We determined zip code–level SES indicators: percentage of persons under poverty in the last 12 months and percentage of adults ages 25 years old or older with a high school degree or higher education. We collected facility-level data: profit status (for profit or nonprofit), volume (defined as the number of patients receiving hemodialysis at the end of the annual survey period), presence and size of a home dialysis program, and staffing. The presence and size of a home dialysis program were ascertained to account for differences in staffing between clinics with and without home programs, because dialysis facilities do not report staff separately for home and in-center dialysis. For staffing, we focused on registered nurses, licensed practical/visiting nurses, and patient care technicians. To determine the total staffing level, we used a previously outlined approach (13) to calculate the number of full–time equivalent positions, considering each reported part-time employee as a 0.5 full-time employee. To derive a composite index of the number of nurses, we added the number of licensed practical/visiting nurses and registered nurses. For facility-level metrics, we assigned each patient the facility reported data on the basis of day 91 of dialysis.
Infection-Related Hospitalization
We examined the rate of hospitalizations for infection during follow-up. We excluded hospitalizations for which the date of admission was the same as the date of discharge, and we combined overlapping hospitalizations, selecting discharge diagnoses from the first or longest hospital record (7% of all hospital records were combined). We classified a hospitalization as infection related if the principal discharge diagnosis included International Classification of Diseases, 9th revision, Clinical Modification (ICD-9-CM) codes of interest (Supplemental Table 1) (14,15). Types of infection examined included dialysis access or central venous catheter (CVC) related; bloodstream infection or sepsis; pulmonary; gastrointestinal, peritoneal, or hepatobiliary; genitourinary; device, procedure, or surgery related (excluding dialysis access); skin and soft tissue; bone and joint; central nervous system; and cardiac.
We followed our cohort for all infection–related hospitalization(s) starting on day 91 of dialysis until the time of death (n=34,778), transplant (n=2338), recovery of kidney function (n=1913), change in dialysis modality (n=5200), change in facility (n=59,537), change in facility profit status (n=522), or study end on December 31, 2009 (n=31,257).
Statistical Analyses
To examine the associations among patient– and facility–level factors and the rate of infection-related hospitalization, we used a Poisson mixed effects model with facility identification as a random intercept using SAS procedure PROC GLIMMIX with Laplace as the approximation method. The total follow-up time for each person (days at risk) excluded periods of hospitalization. In the primary multivariable model, we included patient characteristics (age, sex, race, ethnicity, BMI, coexisting illnesses, vascular access type, nephrology care before dialysis initiation, residence in a care facility, tobacco use, alcohol and/or drug dependence, eGFR, serum albumin, ESRD network, zip code–level SES indicators, and RUCA codes) and facility characteristics (volume, number and composition of staff, and presence and size of a home dialysis program). Volume was classified into quartiles, because we did not have a priori cutpoints. The presence and size of a home dialysis program were classified as none/0, 1–10, 11–25, or ≥26 patients. Two variables were included to account for staffing: (1) total staffing (total number of nurses + patient care technicians) and (2) proportion of staff composed of nurses (nurses/[nurses + patient care technicians]). We examined the rate of all infection hospitalizations, and then, we applied the same Poisson mixed effects model to examine the rates of four specific types of infection-related hospitalization: (1) bloodstream infections or sepsis, (2) dialysis access, (3) pulmonary, or (4) other types of infection. We express rate ratios (risk) derived from the Poisson model comparing rates across patient and facility characteristics. We used multiple imputation to create 10 datasets and imputed the following when missing: albumin (23%), eGFR (0.4%), BMI (1.3%), and dialysis access type (0.4%). We also conducted sensitivity analyses, excluding patients with extreme outlier values for volume or staffing defined as values ≤1st or ≥99th percentile. In secondary models, we added an additional facility characteristic—profit status. All statistical analyses were conducted using SAS version 9.4. Our study was not considered human subjects research by the University of California, Davis Institutional Review Board.
Results
We examined 135,545 Medicare beneficiaries (Table 1). During a median follow-up of 313 days (89–664), 38,475 (28%) patients had at least one infection-related hospitalization: 18% had one, 6% had two, 2% had three, and 2% had four or more hospitalizations. The median time to infection-related hospitalization was 161 days (55–379).
Table 1.
Baseline characteristics of Medicare beneficiaries receiving in-center hemodialysis
| Patient and Facility Characteristics | n=135,545 |
|---|---|
| Patient characteristics | |
| Age, yrs | |
| 18–44 | 13,567 (10) |
| 45–64 | 41,782 (31) |
| 65–74 | 37,402 (28) |
| 75–84 | 33,854 (25) |
| ≥85 | 8940 (7) |
| Men | 73,466 (54) |
| Race | |
| White | 88,115 (65) |
| Black | 41,346 (31) |
| Asian | 3436 (3) |
| Native American | 1505 (1) |
| Pacific Islander | 765 (1) |
| Multiracial, other, unknown | 378 (0.3) |
| Hispanic | 16,011 (12) |
| Body mass index, kg/m2 | |
| <20 | 11,061 (8) |
| 20 to <25 | 37,735 (28) |
| 25 to <30 | 38,419 (28) |
| 30 to <35 | 23,265 (17) |
| ≥35 | 23,352 (17) |
| Missing | 1713 (1) |
| Diabetes mellitus | 80,818 (60) |
| Atherosclerotic heart disease | 32,055 (24) |
| Cerebrovascular disease | 14,847 (11) |
| Peripheral vascular disease | 21,435 (16) |
| Congestive heart failure | 49,910 (37) |
| History of amputation | 4580 (3) |
| Hypertension | 119,756 (88) |
| Cancer | 10,062 (7) |
| COPD | 13,606 (10) |
| Inability to ambulate or transfer | 9827 (7) |
| Etiology of renal disease | |
| Diabetes | 64,106 (47) |
| Hypertension | 42,079 (31) |
| GN | 7350 (5) |
| Cystic kidney | 1927 (1) |
| Other or unknown | 20,083 (15) |
| Access type | |
| AV fistula | 18,710 (14) |
| AV graft | 6349 (5) |
| Catheter | 108,262 (80) |
| Other | 1659 (1) |
| Missing | 565 (0.4) |
| Hemoglobin, g/dl | 10.0±1.7 |
| eGFR, ml/min per 1.73 m2 | 11.1±5.7 |
| Albumin, g/dl | |
| <2.5 | 18,508 (14) |
| 2.5–2.9 | 22,572 (17) |
| 3.0–3.4 | 28,706 (21) |
| 3.5–3.9 | 23,188 (17) |
| ≥4 | 10,949 (8) |
| Missing | 31,622 (23) |
| Tobacco use, current smoker | 8841 (7) |
| Alcohol dependence | 1939 (1) |
| Drug dependence | 1772 (1) |
| Institutionalization | |
| Assisted living | 1000 (1) |
| Nursing home | 9321 (7) |
| Other | 581 (0.4) |
| Prior nephrologist care, mo | |
| None | 41,366 (31) |
| <6 | 14,088 (10) |
| 6–12 | 32,560 (24) |
| >12 | 31,388 (23) |
| Unknown | 16,143 (12) |
| Geographic characteristics | |
| RUCA | |
| Isolated small rural | 5796 (4) |
| Small rural | 8825 (7) |
| Large rural | 15,698 (12) |
| Urban | 105,226 (78) |
| ESRD network (patient) | |
| 1 | 4815 (4) |
| 2 | 5703 (4) |
| 3 | 3671 (3) |
| 4 | 5733 (4) |
| 5 | 8181 (6) |
| 6 | 14,687 (11) |
| 7 | 9034 (7) |
| 8 | 9084 (7) |
| 9 | 11,577 (9) |
| 10 | 5383 (4) |
| 11 | 8321 (6) |
| 12 | 5537 (4) |
| 13 | 7102 (5) |
| 14 | 14,060 (10) |
| 15 | 5482 (4) |
| 16 | 3580 (3) |
| 17 | 4973 (4) |
| 18 | 8622 (6) |
| Socioeconomic status | |
| Adults ≥25 yrs old with high school education or higher, % | |
| <50 | 1782 (1) |
| 50–75 | 27,058 (20) |
| >75 | 106,705 (79) |
| Below poverty in past 12 mo, % | |
| <5 | 9397 (7) |
| 5–9.9 | 25,303 (19) |
| 10–14.9 | 27,746 (20) |
| 15–19.9 | 25,215 (19) |
| 20–24.9 | 18,623 (14) |
| ≥25 | 29,261 (22) |
| Facility characteristics | |
| Volume, no. of facility patients | |
| ≤41 | 18,409 (14) |
| 42–63 | 28,549 (21) |
| 64–91 | 37,254 (27) |
| ≥92 | 51,333 (38) |
| Volume of home program | |
| 0/None | 74,735 (55) |
| 1–10 | 27,796 (21) |
| 11–25 | 20,135 (15) |
| ≥26 | 12,879 (10) |
| Total staffing, nurses, and PCTs | 13 [9, 18] |
| Staffing composition, % nursesa | 41 [33, 50] |
| For profit | 123,697 (91) |
Data are presented as n (%), mean±SD, or median [25th, 75th percentiles]. COPD, chronic obstructive pulmonary disease; AV, arteriovenous; RUCA, rural-urban commuting area; PCT, patient care technician.
(Nurses/[nurses + PCTs]) × 100.
The overall rate of infection-related hospitalization was 40.2 per 100 person-years. The rates of hospitalization were highest for dialysis access or CVC-related infections, bloodstream infections or sepsis, and pulmonary infections, with rates (per 100 person-years) of 11.9, 10.2, and 8.4, respectively.
All–Cause Infection–Related Hospitalization
In the primary multivariable model, patient characteristics associated with a higher rate of infection-related hospitalization included age <45 years old or 75 years old and older, women, white race, non-Hispanic ethnicity, BMI<20 kg/m2, coexisting illnesses, inability to ambulate or transfer, drug dependence, residence in a care facility, higher eGFR, lower serum albumin concentrations at dialysis initiation, dialysis initiation with an access other than an arteriovenous fistula, and limited nephrology care before dialysis (Table 2). Patient residence in an urban area was associated with a higher rate of infection-related hospitalization, whereas measures of neighborhood SES were not, and ESRD networks were strongly associated with the rate of hospitalization for infection (Table 2).
Table 2.
Multivariable model examining characteristics associated with all–cause infection–related hospitalization
| Variable | Rate Ratio (95% Confidence Interval) | P Value |
|---|---|---|
| Age, yrs | ||
| 18–44 | 1.12 (1.07 to 1.17) | <0.001 |
| 45–64 | Reference | |
| 65–74 | 1.03 (1.00 to 1.06) | 0.08 |
| 75–84 | 1.14 (1.10 to 1.17) | <0.001 |
| ≥85 | 1.30 (1.23 to 1.37) | <0.001 |
| Men | 0.86 (0.84 to 0.88) | <0.001 |
| Race | ||
| White | Reference | |
| Black | 0.82 (0.79 to 0.85) | <0.001 |
| Other | 0.85 (0.80 to 0.90) | <0.001 |
| Hispanic | 0.89 (0.85 to 0.93) | <0.001 |
| Body mass index, kg/m2 | ||
| <20 | 1.10 (1.06 to 1.15) | <0.001 |
| 20 to <25 | Reference | |
| 25 to <30 | 0.95 (0.92 to 0.98) | <0.001 |
| 30 to <35 | 0.95 (0.91 to 0.98) | 0.002 |
| ≥35 | 1.00 (0.97 to 1.04) | 0.94 |
| Congestive heart failure | 1.09 (1.06 to 1.11) | <0.001 |
| Atherosclerotic heart disease | 1.01 (0.98 to 1.04) | 0.70 |
| Cerebrovascular disease | 1.12 (1.08 to 1.16) | <0.001 |
| Peripheral vascular disease | 1.10 (1.07 to 1.14) | <0.001 |
| Hypertension | 0.84 (0.81 to 0.87) | <0.001 |
| History of amputation | 1.19 (1.11 to 1.26) | <0.001 |
| Diabetes mellitus | 1.16 (1.13 to 1.19) | <0.001 |
| Cancer | 1.20 (1.15 to 1.25) | <0.001 |
| COPD | 1.25 (1.21 to 1.30) | <0.001 |
| Inability to ambulate or transfer | 1.33 (1.27 to 1.39) | <0.001 |
| Vascular access type | ||
| AV fistula | Reference | |
| Catheter | 1.59 (1.53 to 1.65) | <0.001 |
| AV graft | 1.37 (1.28 to 1.45) | <0.001 |
| Other | 1.46 (1.30 to 1.64) | <0.001 |
| eGFR per 5 ml/min per 1.73 m2 | 1.11 (1.10 to 1.12) | <0.001 |
| Albumin concentration, g/dl | ||
| <2.5 | 1.44 (1.35 to 1.53) | <0.001 |
| 2.5 to <3.0 | 1.31 (1.24 to 1.39) | <0.001 |
| 3.0 to <3.5 | 1.21 (1.14 to 1.29) | <0.001 |
| 3.5 to <4.0 | 1.11 (1.05 to 1.18) | <0.001 |
| ≥4 | Reference | |
| Tobacco use, current smoker | 0.98 (0.94 to 1.03) | 0.51 |
| Alcohol dependence | 1.05 (0.95 to 1.15) | 0.35 |
| Drug dependence | 1.26 (1.14 to 1.40) | <0.001 |
| Residence in care facility | 1.51 (1.44 to 1.58) | <0.001 |
| Nephrology care before dialysis initiation, mo | ||
| <6 | 1.04 (1.00 to 1.09) | 0.07 |
| 6–12 | 1.04 (1.01 to 1.07) | 0.02 |
| ≥12 | Reference | |
| None | 1.13 (1.09 to 1.17) | <0.001 |
| Unknown | 1.19 (1.14 to 1.25) | <0.001 |
| RUCA | ||
| Urban | Reference | |
| Large rural | 0.96 (0.91 to 1.00) | 0.04 |
| Small rural | 0.95 (0.90 to 1.00) | 0.04 |
| Isolated small rural | 0.91 (0.86 to 0.97) | 0.003 |
| ESRD network | ||
| 1 | 1.16 (1.03 to 1.30) | 0.01 |
| 2 | 1.14 (1.01 to 1.29) | 0.04 |
| 3 | 1.28 (1.11 to 1.48) | <0.001 |
| 4 | 1.13 (1.01 to 1.27) | 0.04 |
| 5 | 1.28 (1.15 to 1.42) | <0.001 |
| 6 | 1.25 (1.14 to 1.38) | <0.001 |
| 7 | 1.36 (1.23 to 1.51) | <0.001 |
| 8 | 1.20 (1.08 to 1.33) | <0.001 |
| 9 | 1.22 (1.10 to 1.35) | <0.001 |
| 10 | 1.20 (1.08 to 1.35) | 0.001 |
| 11 | 1.15 (1.04 to 1.28) | <0.01 |
| 12 | 1.13 (1.01 to 1.26) | 0.04 |
| 13 | 1.29 (1.16 to 1.44) | <0.001 |
| 14 | 1.15 (1.04 to 1.27) | <0.01 |
| 15 | 1.10 (0.99 to 1.23) | 0.08 |
| 16 | Reference | |
| 17 | 1.10 (0.99 to 1.23) | 0.09 |
| 18 | 1.18 (1.06 to 1.32) | 0.002 |
| Adults ≥25 yrs old who are high school graduates or higher, % | ||
| <50 | Reference | |
| 50–75 | 1.03 (0.92 to 1.16) | 0.60 |
| >75 | 1.01 (0.89 to 1.15) | 0.85 |
| All people below poverty in the past 12 mo, % | ||
| <5 | Reference | |
| 5–9.9 | 0.98 (0.93 to 1.04) | 0.59 |
| 10–14.9 | 0.99 (0.93 to 1.05) | 0.69 |
| 15–19.9 | 1.00 (0.94 to 1.06) | 0.92 |
| 20–24.9 | 1.02 (0.96 to 1.08) | 0.54 |
| ≥25 | 1.04 (0.97 to 1.10) | 0.27 |
| Volume, no. of facility patients | ||
| ≤41 | Reference | |
| 42–63 | 1.04 (0.99 to 1.09) | 0.12 |
| 64–91 | 1.05 (1.00 to 1.11) | 0.05 |
| >91 | 1.04 (0.97 to 1.11) | 0.29 |
| Total no. of nurses and PCTs per five staff increase | 0.99 (0.98 to 1.01) | 0.29 |
| Staff composition per 10% nurse increase | 0.99 (0.98 to 1.00) | 0.10 |
| Home program volume | ||
| None or 0 | Reference | |
| 1–10 | 1.07 (1.03 to 1.11) | <0.001 |
| 11–25 | 1.02 (0.97 to 1.07) | 0.47 |
| ≥26 | 1.00 (0.94 to 1.06) | 0.98 |
COPD, chronic obstructive pulmonary disease; AV, arteriovenous; RUCA, rural-urban commuting area; PCT, patient care technician.
In the multivariable model, higher facility volume was associated with a higher rate, but only a facility volume of 64–91 patients was statistically significantly higher compared with facilities with <42 patients (P value =0.05). The total number and composition of staff were not associated with overall infection–related hospitalization rates (Table 2). In sensitivity analyses, when we removed outliers, we did not find substantive differences in the associations among patient, geographic, and facility characteristics and all–cause infection–related hospitalization, although nephrology care <6 months became significantly associated with a higher rate.
In secondary analyses, in which profit status was included as an additional facility characteristic, care in a for-profit facility was associated with an 11% higher rate of infection-related hospitalization (Supplemental Table 2). The addition of profit status to the model did not substantively change the interpretation of the other characteristics under consideration. However, select ESRD networks were no longer statistically significantly associated with the rate of infection (Table 2, Supplemental Table 2).
Cause–Specific Infection–Related Hospitalization
The association between patient characteristics and the rate of infection-related hospitalization differed by type of infection. For example, chronic obstructive pulmonary disease was associated with a nearly 80% higher rate of hospitalization for pulmonary infections, whereas dialysis catheters at dialysis initiation were associated with a 3-fold higher rate of hospitalization for dialysis access infection (Table 3).
Table 3.
Multivariable model examining characteristics associated with cause–specific infection–related hospitalization
| Variable | Bloodstream Infection or Sepsis | Dialysis Access Infection | Pulmonary Infection | Other Infection | ||||
|---|---|---|---|---|---|---|---|---|
| RR (95% CI) | P Value | RR (95% CI) | P Value | RR (95% CI) | P Value | RR (95% CI) | P Value | |
| Age, yrs | ||||||||
| 18–44 | 0.86 (0.79 to 0.94) | <0.001 | 1.34 (1.26 to 1.42) | <0.001 | 0.98 (0.90 to 1.07) | 0.62 | 1.07 (0.99 to 1.15) | 0.08 |
| 45–64 | Reference | Reference | Reference | Reference | ||||
| 65–74 | 1.19 (1.13 to 1.26) | <0.001 | 0.84 (0.80 to 0.88) | <0.001 | 1.17 (1.10 to 1.23) | <0.001 | 1.00 (0.95 to 1.06) | 0.97 |
| 75–84 | 1.39 (1.31 to 1.47) | <0.001 | 0.82 (0.78 to 0.87) | <0.001 | 1.36 (1.28 to 1.45) | <0.001 | 1.11 (1.05 to 1.18) | <0.001 |
| ≥85 | 1.57 (1.44 to 1.71) | <0.001 | 0.93 (0.85 to 1.01) | 0.08 | 1.61 (1.48 to 1.75) | <0.001 | 1.26 (1.16 to 1.38) | <0.001 |
| Men | 0.91 (0.87 to 0.95) | <0.001 | 0.90 (0.87 to 0.93) | <0.001 | 0.90 (0.86 to 0.94) | <0.001 | 0.72 (0.69 to 0.75) | <0.001 |
| Race | ||||||||
| White | Reference | Reference | Reference | Reference | ||||
| Black | 0.92 (0.87 to 0.98) | <0.01 | 0.95 (0.90 to 1.00) | 0.05 | 0.68 (0.64 to 0.72) | <0.001 | 0.66 (0.63 to 0.70) | <0.001 |
| Other | 0.89 (0.79 to 1.00) | 0.05 | 0.98 (0.89 to 1.09) | 0.74 | 0.86 (0.77 to 0.95) | 0.004 | 0.68 (0.61 to 0.76) | <0.001 |
| Hispanic | 0.86 (0.79 to 0.93) | <0.001 | 0.92 (0.85 to 0.99) | 0.03 | 0.89 (0.82 to 0.97) | <0.01 | 0.87 (0.81 to 0.94) | <0.001 |
| Body mass index, kg/m2 | ||||||||
| <20 | 1.14 (1.06 to 1.24) | <0.001 | 1.02 (0.95 to 1.10) | 0.53 | 1.20 (1.12 to 1.29) | <0.001 | 1.02 (0.94 to 1.11) | 0.57 |
| 20–24 | Reference | Reference | Reference | Reference | ||||
| 25–29 | 0.97 (0.92 to 1.02) | 0.24 | 0.95 (0.91 to 1.00) | 0.07 | 0.88 (0.83 to 0.92) | <0.001 | 0.98 (0.93 to 1.03) | 0.44 |
| 30–34 | 0.95 (0.89 to 1.01) | 0.10 | 0.95 (0.90 to 1.01) | 0.08 | 0.81 (0.76 to 0.86) | <0.001 | 1.06 (1.00 to 1.13) | 0.05 |
| ≥35 | 1.00 (0.94 to 1.06) | >0.99 | 1.05 (1.00 to 1.11) | 0.07 | 0.75 (0.70 to 0.80) | <0.001 | 1.16 (1.10 to 1.23) | <0.001 |
| Congestive heart failure | 1.13 (1.09 to 1.18) | <0.001 | 1.02 (0.98 to 1.06) | 0.36 | 1.15 (1.10 to 1.21) | <0.001 | 1.07 (1.03 to 1.12) | 0.001 |
| Atherosclerotic heart disease | 0.99 (0.94 to 1.04) | 0.71 | 0.97 (0.93 to 1.02) | 0.29 | 0.99 (0.94 to 1.05) | 0.81 | 1.04 (0.99 to 1.09) | 0.16 |
| Cerebrovascular disease | 1.19 (1.12 to 1.26) | <0.001 | 1.11 (1.05 to 1.17) | <0.001 | 1.07 (1.01 to 1.14) | 0.03 | 1.10 (1.03 to 1.16) | 0.003 |
| Peripheral vascular disease | 1.12 (1.06 to 1.19) | <0.001 | 1.13 (1.07 to 1.19) | <0.001 | 1.01 (0.95 to 1.07) | 0.70 | 1.11 (1.05 to 1.17) | <0.001 |
| Hypertension | 0.75 (0.71 to 0.80) | <0.001 | 0.87 (0.82 to 0.92) | <0.001 | 0.91 (0.85 to 0.97) | 0.003 | 0.84 (0.79 to 0.90) | <0.001 |
| History of amputation | 1.31 (1.18 to 1.45) | <0.001 | 1.15 (1.05 to 1.26) | 0.004 | 1.04 (0.92 to 1.18) | 0.53 | 1.21 (1.09 to 1.34) | <0.001 |
| Diabetes mellitus | 1.24 (1.18 to 1.30) | <0.001 | 1.11 (1.06 to 1.16) | <0.001 | 1.03 (0.99 to 1.08) | 0.15 | 1.28 (1.22 to 1.34) | <0.001 |
| COPD | 1.20 (1.13 to 1.29) | <0.001 | 1.10 (1.03 to 1.18) | 0.003 | 1.78 (1.68 to 1.89) | <0.001 | 1.05 (0.98 to 1.12) | 0.17 |
| Cancer | 1.28 (1.19 to 1.38) | <0.001 | 1.11 (1.03 to 1.20) | <0.01 | 1.19 (1.10 to 1.28) | <0.001 | 1.18 (1.09 to 1.28) | <0.001 |
| Inability to ambulate or transfer | 1.57 (1.46 to 1.69) | <0.001 | 1.21 (1.12 to 1.31) | <0.001 | 1.11 (1.02 to 1.22) | 0.02 | 1.32 (1.22 to 1.42) | <0.001 |
| Vascular access type | ||||||||
| AV fistula | Reference | Reference | Reference | Reference | ||||
| Catheter | 1.54 (1.44 to 1.65) | <0.001 | 3.07 (2.82 to 3.36) | <0.001 | 1.14 (1.07 to 1.22) | <0.001 | 1.32 (1.24 to 1.40) | <0.001 |
| AV graft | 1.41 (1.28 to 1.57) | <0.001 | 2.36 (2.09 to 2.68) | <0.001 | 1.06 (0.95 to 1.18) | 0.30 | 1.15 (1.04 to 1.28) | <0.01 |
| Other | 1.43 (1.18 to 1.73) | <0.001 | 2.93 (2.39 to 3.60) | <0.001 | 1.04 (0.85 to 1.26) | 0.71 | 1.17 (0.95 to 1.45) | 0.15 |
| eGFR per 5 ml/min per 1.73 m2 | 1.15 (1.13 to 1.17) | <0.001 | 1.08 (1.07 to 1.10) | <0.001 | 1.09 (1.07 to 1.11) | <0.001 | 1.10 (1.09 to 1.12) | <0.001 |
| Albumin concentration, g/dl | ||||||||
| <2.5 | 1.54 (1.38 to 1.72) | <0.001 | 1.36 (1.25 to 1.49) | <0.001 | 1.48 (1.33 to 1.65) | <0.001 | 1.38 (1.25 to 1.51) | <0.001 |
| 2.5 to < 3.0 | 1.36 (1.23 to 1.50) | <0.001 | 1.28 (1.18 to 1.40) | <0.001 | 1.32 (1.18 to 1.46) | <0.001 | 1.28 (1.17 to 1.41) | <0.001 |
| 3.0 to < 3.5 | 1.27 (1.14 to 1.41) | <0.001 | 1.15 (1.05 to 1.25) | 0.002 | 1.28 (1.17 to 1.40) | <0.001 | 1.17 (1.08 to 1.28) | <0.001 |
| 3.5 to < 4.0 | 1.08 (0.98 to 1.20) | 0.13 | 1.11 (1.02 to 1.21) | 0.02 | 1.21 (1.10 to 1.33) | <0.001 | 1.05 (0.96 to 1.16) | 0.27 |
| ≥4 | Reference | Reference | Reference | Reference | ||||
| Tobacco use, current smoker | 0.91 (0.83 to 0.99) | 0.02 | 0.96 (0.89 to 1.04) | 0.28 | 1.15 (1.06 to 1.25) | <0.001 | 0.91 (0.84 to 0.99) | 0.03 |
| Alcohol dependence | 1.02 (0.85 to 1.21) | 0.86 | 1.05 (0.90 to 1.22) | 0.54 | 0.84 (0.70 to 1.01) | 0.06 | 1.24 (1.04 to 1.48) | 0.02 |
| Drug dependence | 1.14 (0.93 to 1.40) | 0.21 | 1.20 (1.03 to 1.40) | 0.02 | 1.45 (1.20 to 1.75) | <0.001 | 1.30 (1.08 to 1.57) | <0.01 |
| Residence in care facility | 1.76 (1.64 to 1.89) | <0.001 | 1.65 (1.54 to 1.76) | <0.001 | 1.15 (1.06 to 1.25) | <0.001 | 1.43 (1.32 to 1.53) | <0.001 |
| Nephrology care before dialysis initiation, mo | ||||||||
| <6 | 1.08 (1.00 to 1.17) | 0.06 | 1.07 (1.00 to 1.16) | 0.06 | 1.04 (0.96 to 1.12) | 0.35 | 1.02 (0.95 to 1.10) | 0.55 |
| 6–12 | 1.09 (1.02 to 1.16) | <0.01 | 1.07 (1.01 to 1.13) | 0.03 | 1.05 (0.99 to 1.11) | 0.13 | 1.00 (0.95 to 1.06) | 0.92 |
| ≥12 | Reference | Reference | Reference | Reference | ||||
| None | 1.24 (1.17 to 1.33) | <0.001 | 1.21 (1.14 to 1.28) | <0.001 | 1.03 (0.97 to 1.10) | 0.30 | 1.04 (0.98 to 1.10) | 0.18 |
| Unknown | 1.30 (1.20 to 1.40) | <0.001 | 1.25 (1.16 to 1.34) | <0.001 | 1.07 (0.99 to 1.16) | 0.09 | 1.21 (1.12 to 1.31) | <0.001 |
| RUCA | ||||||||
| Urban | Reference | Reference | Reference | Reference | ||||
| Large rural | 1.01 (0.94 to 1.09) | 0.76 | 0.89 (0.83 to 0.95) | <0.001 | 1.02 (0.95 to 1.09) | 0.62 | 0.91 (0.85 to 0.97) | <0.01 |
| Small rural | 0.95 (0.86 to 1.04) | 0.24 | 0.91 (0.83 to 1.00) | 0.04 | 1.06 (0.97 to 1.16) | 0.18 | 0.87 (0.79 to 0.95) | 0.003 |
| Isolated small rural | 0.93 (0.83 to 1.03) | 0.15 | 0.92 (0.83 to 1.02) | 0.10 | 1.01 (0.91 to 1.12) | 0.88 | 0.81 (0.73 to 0.90) | <0.001 |
| ESRD network | ||||||||
| 1 | 0.78 (0.64 to 0.95) | 0.01 | 1.13 (0.93 to 1.38) | 0.21 | 1.36 (1.11 to 1.65) | 0.002 | 1.46 (1.22 to 1.76) | <0.001 |
| 2 | 1.16 (0.94 to 1.43) | 0.16 | 1.22 (1.01 to 1.47) | 0.04 | 1.00 (0.82 to 1.23) | 0.97 | 1.17 (0.97 to 1.41) | 0.10 |
| 3 | 1.10 (0.88 to 1.39) | 0.40 | 1.37 (1.11 to 1.68) | 0.003 | 1.13 (0.89 to 1.42) | 0.32 | 1.44 (1.19 to 1.76) | <0.001 |
| 4 | 1.05 (0.87 to 1.28) | 0.60 | 1.26 (1.04 to 1.53) | 0.02 | 1.11 (0.91 to 1.35) | 0.30 | 1.15 (0.97 to 1.37) | 0.12 |
| 5 | 1.19 (0.99 to 1.43) | 0.06 | 1.35 (1.13 to 1.60) | <0.001 | 1.34 (1.11 to 1.61) | 0.002 | 1.27 (1.07 to 1.51) | <0.01 |
| 6 | 1.07 (0.90 to 1.28) | 0.41 | 1.47 (1.25 to 1.73) | <0.001 | 1.30 (1.09 to 1.55) | 0.004 | 1.18 (1.00 to 1.38) | 0.05 |
| 7 | 1.15 (0.97 to 1.37) | 0.11 | 1.85 (1.56 to 2.19) | <0.001 | 1.01 (0.84 to 1.22) | 0.91 | 1.39 (1.18 to 1.64) | <0.001 |
| 8 | 1.02 (0.85 to 1.22) | 0.83 | 1.30 (1.09 to 1.54) | 0.003 | 1.29 (1.07 to 1.56) | <0.01 | 1.23 (1.04 to 1.46) | 0.02 |
| 9 | 1.05 (0.88 to 1.25) | 0.58 | 1.26 (1.06 to 1.50) | <0.01 | 1.33 (1.11 to 1.59) | 0.002 | 1.22 (1.04 to 1.44) | 0.02 |
| 10 | 1.08 (0.89 to 1.30) | 0.45 | 1.13 (0.93 to 1.36) | 0.22 | 1.39 (1.15 to 1.69) | <0.001 | 1.32 (1.10 to 1.58) | 0.003 |
| 11 | 1.06 (0.89 to 1.27) | 0.52 | 1.36 (1.14 to 1.62) | <0.001 | 1.07 (0.89 to 1.30) | 0.47 | 1.10 (0.93 to 1.31) | 0.27 |
| 12 | 0.86 (0.70 to 1.04) | 0.12 | 1.19 (0.99 to 1.43) | 0.07 | 1.32 (1.09 to 1.62) | <0.01 | 1.23 (1.03 to 1.47) | 0.02 |
| 13 | 1.18 (0.98 to 1.42) | 0.08 | 1.49 (1.25 to 1.77) | <0.001 | 1.33 (1.10 to 1.62) | 0.003 | 1.18 (0.99 to 1.41) | 0.06 |
| 14 | 0.96 (0.81 to 1.15) | 0.67 | 1.24 (1.06 to 1.47) | <0.01 | 1.16 (0.97 to 1.39) | 0.11 | 1.19 (1.01 to 1.39) | 0.04 |
| 15 | 0.87 (0.71 to 1.05) | 0.15 | 1.24 (1.03 to 1.50) | 0.02 | 1.21 (1.00 to 1.47) | 0.05 | 1.11 (0.92 to 1.34) | 0.26 |
| 16 | Reference | Reference | Reference | Reference | ||||
| 17 | 1.09 (0.90 to 1.31) | 0.39 | 1.01 (0.83 to 1.22) | 0.93 | 1.24 (1.02 to 1.50) | 0.03 | 1.09 (0.90 to 1.31) | 0.38 |
| 18 | 1.37 (1.14 to 1.64) | <0.001 | 1.13 (0.95 to 1.34) | 0.18 | 1.07 (0.88 to 1.30) | 0.50 | 1.10 (0.92 to 1.31) | 0.28 |
| Adults ≥25 yrs old who are high school graduates or higher, % | ||||||||
| <50 | Reference | Reference | Reference | Reference | ||||
| 50–75 | 0.95 (0.79 to 1.13) | 0.55 | 1.17 (0.98 to 1.41) | 0.09 | 0.93 (0.77 to 1.13) | 0.48 | 1.04 (0.88 to 1.23) | 0.66 |
| >75 | 0.89 (0.74 to 1.08) | 0.24 | 1.17 (0.97 to 1.41) | 0.11 | 0.90 (0.74 to 1.10) | 0.30 | 1.04 (0.87 to 1.24) | 0.69 |
| All people below poverty in the past 12 mo, % | ||||||||
| <5 | Reference | Reference | Reference | Reference | ||||
| 5–9.9 | 0.97 (0.88 to 1.07) | 0.51 | 0.98 (0.90 to 1.08) | 0.70 | 1.03 (0.94 to 1.13) | 0.55 | 0.97 (0.89 to 1.06) | 0.53 |
| 10–14.9 | 0.96 (0.87 to 1.06) | 0.47 | 0.99 (0.90 to 1.08) | 0.76 | 1.03 (0.94 to 1.13) | 0.55 | 0.98 (0.89 to 1.06) | 0.58 |
| 15–19.9 | 0.97 (0.88 to 1.07) | 0.54 | 0.99 (0.90 to 1.09) | 0.84 | 1.05 (0.95 to 1.15) | 0.36 | 1.02 (0.93 to 1.12) | 0.70 |
| 20–24.9 | 0.99 (0.89 to 1.10) | 0.90 | 0.99 (0.90 to 1.09) | 0.86 | 1.12 (1.01 to 1.25) | 0.03 | 1.01 (0.91 to 1.11) | 0.85 |
| ≥25 | 0.98 (0.88 to 1.10) | 0.77 | 1.06 (0.96 to 1.18) | 0.23 | 1.10 (0.99 to 1.22) | 0.09 | 1.02 (0.93 to 1.13) | 0.65 |
| Volume, no. of facility patients | ||||||||
| ≤41 | Reference | Reference | Reference | Reference | ||||
| 42–63 | 1.03 (0.95 to 1.13) | 0.47 | 1.02 (0.94 to 1.10) | 0.65 | 1.03 (0.95 to 1.11) | 0.48 | 1.09 (1.01 to 1.17) | 0.03 |
| 64–91 | 1.02 (0.93 to 1.12) | 0.68 | 1.05 (0.97 to 1.14) | 0.23 | 1.12 (1.03 to 1.22) | <0.01 | 1.07 (0.99 to 1.16) | 0.07 |
| >91 | 0.96 (0.86 to 1.08) | 0.50 | 1.06 (0.96 to 1.18) | 0.22 | 1.16 (1.05 to 1.28) | 0.003 | 1.08 (0.99 to 1.19) | 0.10 |
| Total no. of nurses and PCTs per five staff higher | 1.00 (0.98 to 1.03) | 0.83 | 0.98 (0.96 to 1.01) | 0.20 | 0.97 (0.95 to 0.99) | 0.01 | 0.98 (0.96 to 1.00) | 0.12 |
| Staff composition per 10% higher nurse composition | 1.00 (0.98 to 1.01) | 0.72 | 0.98 (0.96 to 0.99) | 0.004 | 1.00 (0.98 to 1.02) | 0.97 | 1.00 (0.98 to 1.02) | 0.98 |
| Home program volume | ||||||||
| 0/None | Reference | Reference | Reference | Reference | ||||
| 1–10 | 1.04 (0.98 to 1.11) | 0.18 | 1.06 (1.00 to 1.13) | 0.04 | 1.07 (1.01 to 1.13) | 0.03 | 1.07 (1.01 to 1.14) | 0.02 |
| 11–25 | 0.99 (0.92 to 1.07) | 0.84 | 1.03 (0.95 to 1.11) | 0.48 | 1.03 (0.96 to 1.10) | 0.45 | 1.04 (0.98 to 1.11) | 0.20 |
| ≥26 | 0.98 (0.88 to 1.08) | 0.66 | 1.00 (0.91 to 1.10) | 0.95 | 0.93 (0.84 to 1.01) | 0.10 | 1.01 (0.92 to 1.10) | 0.82 |
RR, rate ratio; 95% CI, 95% confidence interval; COPD, chronic obstructive pulmonary disease; AV, arteriovenous; RUCA, rural-urban commuting area; HS, high school; PCT, patient care technician.
Rates of hospitalization for dialysis access–related infection varied substantially across ESRD networks. Rural-urban residence was significantly associated with hospitalization for dialysis access and other types of infection. SES, as measured by poverty and educational attainment at the zip code level, did not have consistent associations with the rate of cause–specific infection–related hospitalization (Table 3).
Interestingly, dialysis facility volume was associated with hospitalization for pulmonary infections. A higher percentage of nurses was associated with a lower rate of hospitalization for dialysis access infections, whereas a higher total number of staff was associated with a lower rate of hospitalization for pulmonary infections. In sensitivity analyses, when we removed outliers, we did not find substantive differences in the effect size of associations between patient, geographic, and facility characteristics and the rate of cause–specific infection–related hospitalization. However, some marginally nonsignificant P values became significant, whereas some marginally significant P values became nonsignificant. When profit status was added to the primary model, for-profit status was only associated with a higher rate of hospitalization for dialysis access or CVC-related infection (Supplemental Table 2). The addition of profit status to the models examining cause-specific infection did not substantively change the interpretation of the vast majority of other characteristics under consideration. However, some P values became significant, whereas others became nonsignificant (Table 3, Supplemental Table 2).
Discussion
We systematically examined the associations among patient, geographic, and facility characteristics, including indicators of SES, and the rates of overall and cause–specific infection–related hospitalizations among Medicare beneficiaries starting in-center hemodialysis. We undertook this study to improve our understanding of the relative association of each of these factors with the rate of infection-related hospitalization and inform our understanding of where preventive efforts and intervention may yield the greatest benefit.
Patient characteristics most strongly associated with a higher rate of all–cause infection–related hospitalization included initiation of dialysis with a catheter, albumin concentrations <2.5 g/dl, inability to ambulate or transfer, and residence in a care facility at initiation of dialysis. Although these findings are not surprising, they remind us that we can risk stratify our patients to identify those who are at highest risk. Lack of nephrology care before dialysis initiation was associated with higher rates of bloodstream infections or sepsis and dialysis access–related infections. Despite limiting our examination of hospitalizations to those occurring after the first 90 days, we found an association between nephrology care before dialysis initiation and infection, raising important questions about when nephrologists, dialysis facilities, and others should become accountable for the outcomes of patients who may be relatively newly under their care.
We noted significant variation in the rate of dialysis access infection–related hospitalization by ESRD network, raising the possibility that patients in these regions differ in ways that we have not measured (e.g., access to transplantation and/or other clinical risk factors), facilities themselves have adopted practices that differ region by region, and/or some local network–driven quality initiatives have been effective at lowering the risk of infection. ESRD networks are nonprofit organizations that contract with the Centers for Medicare & Medicaid Services to oversee the quality of care within dialysis facilities (16), and the structure of these networks allows for unique opportunities to provide guidance on preventing infection-related hospitalizations and monitor individual facility performance. Residence in an urban area was associated with higher rates of specific types of infection-related hospitalizations (e.g., dialysis access). Whether this finding reflects access to hospitals, differences in management of infection in urban versus rural outpatient dialysis facilities, or other factors is unknown and may warrant additional exploration.
We did not find consistent associations between measures of poverty or education at the patient zip code level and the risk of hospitalization for infection. The lack of association may have been related to our cohort being restricted to Medicare Parts A and B beneficiaries. Interestingly, another and perhaps better potential measure of access to care—nephrology care before dialysis—was associated with the rate of infection-related hospitalization.
We examined modifiable and nonmodifiable facility–level factors, including volume, staffing, and profit status. Higher volume facilities had higher and graded relative rates of pulmonary infection; higher total staffing was associated with a lower rate of hospitalization for pulmonary infections, whereas a higher percentage of nurses was associated with a lower relative rate of hospitalization for dialysis access–related infections. Secondary models that further included profit status found for-profit status to be associated with a higher rate of overall infections and in analyses of cause-specific hospitalizations, infection related to the dialysis access. These findings highlight our need to better understand how organizational structure, facility characteristics, and staffing influence the risk of infection-related hospitalization. For example, higher facility volume may heighten the risk of infection because of a greater likelihood of exposure to ill patients at the dialysis facility (e.g., viral pulmonary infections) or less attention to infection prevention because of demands on efficiency. The association between higher staffing and lower rate of pulmonary infection may reflect the greater availability of resources needed to counsel patients about vaccination and ensure implementation of infection prevention and control guidelines, whereas a higher percentage of nurses may allow for effective vascular access management. We note that, in the years after our cohort began dialysis therapy, numerous initiatives have been undertaken by dialysis organizations to lower the prevalence of dialysis catheters (17) and the incidence of bloodstream infections (18). Future studies should examine whether and to what extent these initiatives have altered the trajectory of infection-related hospitalizations.
Our study has a number of strengths, including our examination of a broad range of factors that contribute to infection risk, and provides targets for interventions. We examined all–cause and cause–specific infection to better identify and conceptualize risk and opportunities for targeted prevention. We also examined this question within existing frameworks for quality improvement, namely with respect to ESRD networks. Lastly, we examined a problem of great importance—infection—that has received relatively little study.
Our study also has limitations. Our findings are limited to Medicare beneficiaries receiving care in free-standing facilities and those who survive the first 90 days. Our reliance on ICD-9-CM codes is a limitation, and we acknowledge the potential for misclassification of the type of infection. However, the principal inpatient diagnosis is clearly and consistently defined as the diagnosis established as having been principally responsible for the admission of the patient to the hospital. The available staffing data may not accurately reflect true full–time equivalents and presumably, do not include per diem employees; also, the data do not account for skill or training, which may be more important than the absolute number of staff. The SES data were at the zip code level, which may have resulted in misclassification.
Reliance on the Medical Evidence Form to collect data was another limitation of our study. First, our conclusions are limited to patient characteristics at the time of dialysis initiation. Second, we cannot exclude residual confounding. A study conducted between 1995 and 1998 examined the sensitivity and specificity of a prior version of the Medical Evidence Form for comorbidity ascertainment of 17 conditions and found the overall sensitivity to be 0.59 and the specificity to be ≥0.91 (19). In addition, Kim et al. (20) examined the concordance between nephrology care reported on the Medical Evidence Form and the first outpatient Medicare claim for nephrology consultation and found an overall accuracy of 0.70 (the accuracy differed by patient characteristics). Thus, we cannot make definitive conclusions about the association between pre–ESRD nephrology care and the risk of infection-related hospitalization. However, our findings suggest that there may be an important link that warrants additional study.
The findings from our study and the most recent statistics from the 2014 USRDS Annual Data Report (1) remain sobering and a reminder of the burden of serious infections in the dialysis population. Increasing attention has been focused on infection prevention and control as exemplified by a number of initiatives and changes in policy, including the CDC’s NHSN dialysis event reporting system, the CDC’s Dialysis Bloodstream Infection Collaborative (21), select ESRD network initiatives, large dialysis organization quality initiatives targeting reduction of dialysis catheters or bloodstream infections (17,18), and the expanded and defined role of dialysis facility medical directors in infection prevention and control (22,23). The CDC has developed core interventions for dialysis facilities to prevent bloodstream infections among patients on dialysis (24).
In conclusion, our findings can be used to identify patients at higher risk for infection and inform the design of strategies aimed to decrease risk. Age, select comorbidities, inability to ambulate or transfer, and residence in a care facility at dialysis initiation can be used to risk-stratify patients, whereas ongoing facility and physician efforts to reduce tunneled dialysis catheter use are imperative to lowering the risk of infection. Finally, variation across ESRD networks should motivate us to explore whether ESRD networks with the lowest rates of infection-related hospitalization have instituted region–wide infection control initiatives that have proved successful.
Disclosures
L.S.D., B.G., and G.A.K. receive research support from Dialysis Clinic, Inc., Nashville, Tennessee. G.M.C. serves on the Board of Directors of Satellite Healthcare, Inc., San Jose, California. G.A.K. is a consultant for the Renal Research Institute, New York, New York, which is a subsidiary of Fresenius, Waltham, Massachusetts.
Supplementary Material
Acknowledgments
This publication was made possible by grants K23 DK093584 (to L.S.D.), R01 DK092232 (to D.V.N.), K24 DK085446 (to G.M.C.), and K24 DK085153 (to K.L.J.) from the National Institutes of Diabetes and Digestive and Kidney Diseases; grants UL1 TR000002 (to L.S.D. and D.V.N) and UL1 TR000153 (to D.V.N.) from the National Center for Advancing Translational Sciences; and a research grant from Dialysis Clinic, Inc.
The funding sources were not involved in the design or conduct of the study, the collection and management of the data, data analysis, interpretation, preparation, review, or approval of the manuscript. The data reported here have been supplied by the US Renal Data System. The interpretation and reporting of these data are the responsibility of the author(s) and in no way should be seen as an official policy or interpretation of the US Government.
Because Dr. Johansen is a Deputy Editor of CJASN, she was not involved in the peer-review process for this manuscript. Another editor oversaw the peer-review and decision-making process for this manuscript.
Footnotes
Published online ahead of print. Publication date available at www.cjasn.org.
See related editorial, “Infections Requiring Hospitalization in Patients on Hemodialysis,” on pages 2101–2103.
This article contains supplemental material online at http://cjasn.asnjournals.org/lookup/suppl/doi:10.2215/CJN.03050315/-/DCSupplemental.
References
- 1.US Renal Data System : USRDS 2014 Annual Data Report: Atlas of Chronic Kidney Disease and End-Stage Renal Disease in the United States, Bethesda, MD, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, 2014 [Google Scholar]
- 2.Dalrymple LS, Johansen KL, Chertow GM, Cheng SC, Grimes B, Gold EB, Kaysen GA: Infection-related hospitalizations in older patients with ESRD. Am J Kidney Dis 56: 522–530, 2010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Chavers BM, Solid CA, Gilbertson DT, Collins AJ: Infection-related hospitalization rates in pediatric versus adult patients with end-stage renal disease in the United States. J Am Soc Nephrol 18: 952–959, 2007 [DOI] [PubMed] [Google Scholar]
- 4.Dalrymple LS, Mu Y, Romano PS, Nguyen DV, Chertow GM, Delgado C, Grimes B, Kaysen GA, Johansen KL: Outcomes of infection-related hospitalization in Medicare beneficiaries receiving in-center hemodialysis. Am J Kidney Dis 65: 754–762, 2015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Centers for Medicare and Medicaid Services Center for Clinical Standards and Quality: ESRD QIP Payment Year 2014 Program Details, 2013. Available at: http://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/ESRDQIP/. Accessed April 8, 2014
- 6.Centers for Medicare and Medicaid Services Center for Clinical Standards and Quality: ESRD QIP Summary: Payment Years 2012–2016. Available at: http://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/ESRDQIP/Downloads/ESRD-QIP-Sumary-PY2012-16pdf. Accessed November 3, 2014
- 7.Dalrymple LS, Johansen KL, Romano PS, Chertow GM, Mu Y, Ishida JH, Grimes B, Kaysen GA, Nguyen DV: Comparison of hospitalization rates among for-profit and nonprofit dialysis facilities. Clin J Am Soc Nephrol 9: 73–81, 2014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.United States Census Bureau: American Fact Finder. Available at: http://factfinder2.census.gov/faces/nav/jsf/pages/index.xhtml. Accessed August 30, 2013
- 9.ZIP Code to ZCTA Crosswalk, 2009. Available at: http://www.dartmouthatlas.org. Accessed August 30, 2013
- 10.RUCA: 2006 ZIP Version 2.0 Codes. Available at: http://depts.washington.edu/uwruca/ruca-download.php. Accessed March 27, 2013
- 11.KDOQI Clinical Practice Guidelines for Chronic Kidney Disease: Evaluation, Classification, and Stratification. Part 5. Evaluation of laboratory measurements for clinical assessment of kidney disease. Am J Kidney Dis 39: S76–S110, 2002 [PubMed] [Google Scholar]
- 12.RUCA: Using RUCA Data. Available at: http://depts.washington.edu/uwruca/ruca-uses.php. Accessed April 9, 2014
- 13.Yoder LA, Xin W, Norris KC, Yan G: Patient care staffing levels and facility characteristics in U.S. hemodialysis facilities. Am J Kidney Dis 62: 1130–1140, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Hart AC, Stegman MS, Ford B, editors: ICD-9-CM for Physicians, Vol 1 & 2, San Francisco, California, OptumInsight, 2011 [Google Scholar]
- 15.Centers for Disease Control and Prevention: Conversion Table of New ICD-9-CM Codes, 2011. Available at: http://www.cdc.gov/nchs/icd/icd9cm_addenda_guidelines.htm. Accessed July 31, 2012
- 16.ESRD Network Organizations: Available at: http://www.cms.gov/Medicare/End-Stage-Renal-Disease/ESRDNetworkOrganizations/index.html?redirect=/ESRDNetworkOrganizations/. Accessed March 12, 2014
- 17.Wilson SM, Mayne TJ, Krishnan M, Holland J, Volz A, Good LS, Nissenson AR: CathAway fistula vascular access program achieves improved outcomes and sets a new standard of treatment for end-stage renal disease. Hemodial Int 17: 86–93, 2013 [DOI] [PubMed] [Google Scholar]
- 18.Rosenblum A, Wang W, Ball LK, Latham C, Maddux FW, Lacson E, Jr.: Hemodialysis catheter care strategies: A cluster-randomized quality improvement initiative. Am J Kidney Dis 63: 259–267, 2014 [DOI] [PubMed] [Google Scholar]
- 19.Longenecker JC, Coresh J, Klag MJ, Levey AS, Martin AA, Fink NE, Powe NR: Validation of comorbid conditions on the end-stage renal disease medical evidence report: The CHOICE study. Choices for Healthy Outcomes in Caring for ESRD. J Am Soc Nephrol 11: 520–529, 2000 [DOI] [PubMed] [Google Scholar]
- 20.Kim JP, Desai M, Chertow GM, Winkelmayer WC: Validation of reported predialysis nephrology care of older patients initiating dialysis. J Am Soc Nephrol 23: 1078–1085, 2012 [DOI] [PubMed] [Google Scholar]
- 21.Patel PR, Yi SH, Booth S, Bren V, Downham G, Hess S, Kelley K, Lincoln M, Morrissette K, Lindberg C, Jernigan JA, Kallen AJ: Bloodstream infection rates in outpatient hemodialysis facilities participating in a collaborative prevention effort: A quality improvement report. Am J Kidney Dis 62: 322–330, 2013 [DOI] [PubMed] [Google Scholar]
- 22.Centers for Medicare & Medicaid Services (CMS): HHS : Medicare and Medicaid Programs; Conditions for coverage for end-stage renal disease facilities. Final rule. Fed Regist 73: 20369–20484, 2008 [PubMed] [Google Scholar]
- 23.Medical Director Toolkit: Developed by the Forum of ESRD Networks' Medical Advisory Council, 2012. Available at: http://www.fmqai.com/library/attachment-library/Medical%20Director%20Toolkit%20062112[1].pdf. Accessed April 8, 2014 [Google Scholar]
- 24.Center for Disease Control: CDC Approach to BSI Prevention in Dialysis Facilities. Available at: http://www.cdc.gov/dialysis/PDFs/Dialysis-Core-Interventions-5_10_13.pdf. Accessed April 11, 2014
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.