Catheter Infections as a Factor in the Transfer of Continuous Ambulatory Peritoneal Dialysis Patients to Hemodialysis

ORIGINAL INVESTIGATIONS

Catheter Infections as a Factor in the Transfer of Continuous Ambulatory Peritoneal Dialysis Patients to Hemodialysis Beth Piraino, MD, Judith Bernardini, BSN, and Michael Sorkin, MD • The effect of peritoneal catheter infections on the transfer of continuous ambulatory peritoneal dialysis (CAPO) patients to hemodialysis over a 9-year period were examined. Twenty-seven percent (68/247) of all patients were transferred permanently to hemodialysis after a mean of 15 + 14 months of CAPO. An additional 29 0/0 transferred temporarily one or more times during the study period (mean time of peritoneal dialysis, 35 + 23 months). The reasons for permanent transfer to hemodialysis were catheter infections (15/68, 220/0), peritonitis (13/68, 190/0), catheter infections associated with peritonitis (10/68, 150/0), patient preference (9/68, 130/0), mechanical problems (4/68, 6%), noncompliance (7/68, 100/0), inadequate clearance or ultrafiltration (6/68, 90/0), with other reasons for the remainder (4/68, 60/0). Temporary transfers to hemodialysis were also mainly due to catheter infections (32 % ), peritonitis (23 0/0), and simultaneous catheter infections and peritonitis (24 0/0). Catheter infection rates were much higher in the groups that permanently and temporarily were transferred to hemodialysis in comparison with those patients who remained on peritoneal dialysis. We conclude that catheter infections are a leading cause of both temporary and permanent transfer of CAPO patients to hemodialysis. © 1989 by the National Kidney Foundation, Inc. INDEX WORDS: Peritoneal dialysis; hemodialysis; catheter infections; peritonitis.

C

ONTINUOUS ambulatory peritoneal dialysis (CAPO) is an increasingly popular mode of dialysis. However, a large percentage of patients who begin CAPO eventually transfer to hemodialysis. Previously published reasons for such transfer include patient preference, ultrafiltration failure, inadequate clearance, noncompliance, and peritonitis. 1-6 Peritonitis is listed most often as the major cause of CAPO dropout. 3-5 We previously found that patients in our CAPO population with a history of catheter infections were more likely to transfer to hemodialysis than patients who never had a catheter infection, independent of peritonitis history.7 Catheter infections appeared to be an important but seldom recognized cause of CAPO failure. In the present report, we investigated the relationship of catheter infections to the transfer of patients from CAPO to hemodialysis over a 9-year period. METHODS We studied a cohort of 247 patients who started receiving continuous peritoneal dialysis treatment between the beginning of our program in 1979 and April 15, 1987. Data collection was stopped on August 15, 1988. Six patients were undergoing continuous cycler peritoneal dialysis (CCPD); the rest were undergoing CAPD. All information on these patients was entered into a computer data handling system consisting of the MUMPS File Manager and a Digital Equipment Corp (DEC, Pittsburgh, PA) PDP 11144. Data, collected prospectively since 1982, included demographic information (cause of renal failure, age, race, sex, and presence of diabetes), date of insertion of the peritoneal catheter, date of initiation of continuous peritoneal dialysis, and all time spent off peritoneal dialysis.

Forty-eight of 295 patients undergoing peritoneal dialysis during this time were excluded from the present analysis. Exclusions were for transfer to another center (nine patients), transfer to intermittent peritoneal dialysis before April 15, 1987 (20 patients), and age < 18 years (19 patients). Data on the remaining 247 patients were analyzed. Cumulative time of continuous peritoneal dialysis for all 247 patients was 444 years as of August 15, 1988. The mean time of peritoneal dialysis per patient was 22 months (range, five days to 7.5 years). End points for continuous peritoneal dialysis were death (74 patients), transplantation (52 patients), transfer to intermittent peritoneal dialysis after April 15, 1987 (seven patients), recovery of renal function (one patient), and permanent transfer to hemodialysis (68 patients). Forty-five patients remained on CAPD at the time the study ended. We prospectively monitored all infections related to peritoneal dialysis, including exit-site infections, tunnel infections, and peritonitis. Peritonitis was defined as cloudy dialysate with > 100 WBC/p.L and >50% of the WBC being polymorphonuclear cells. Exit-site infections were defined as erythema or drainage or both from the exit site. Thnnel infections were defined as erythema, edema, and tenderness over the subcutaneous peritoneal catheter pathway. Exit -site and tunnel infections were combined for the purpose of this study and referred to as catheter infections. Catheter infections and peritonitis episodes were considered to be associated infections if the patient had both simultaneously, or if the peritonitis developed within 2 From the Renal Electrolyte Division, Department of Medicine, University of Pittsburgh, and the Veterans Administration Medical Center, Pittsburgh. Presented in part at the Eighth National Conference on CAPD, February 10, 1988, Kansas City, MO. Address reprint requests to Beth Piraino, MD, the Renal Electrolyte Division, School of Medicine, University of Pittsburgh, 1191 Scaife Hall, Pittsburgh, PA 15261. © 1989 by the National Kidney Foundation, Inc. 0272-6386/89/1305-0002$3.00/0

American Journal of Kidney Diseases, Vol XIII, No 5 (May), 1989: pp 365-369

365

366

PIRAINO, BERNARDINI, AND SORKIN

weeks of cessation of antibiotics for the catheter infection. In addition, to meet the criteria of association, the same organism had to be present in the dialysate and the catheter exit site, or one site was sterile. The overall peritonitis rate was 0.9 episodes/year and catheter infection rate was 1.1 episode/year for all patients studied. Infectious reasons for removal of a peritoneal" catheter were (1) intractable peritonitis that did not resolve with appropriate antibiotic therapy based on dialysate culture and sensitivity results, (2) persistent catheter infections that did not resolve after multiple courses of antibiotics, and (3) catheter infections associated with peritonitis that did not respond to antibiotic therapy. An average of three courses (ten to 14 days each) of antibiotics were given before removal of the catheter. Antibiotic therapy ranged from one day (for severely ill patients) to 12 courses (for a patient with a persistent exit-site infection). We recorded temporary and permanent transfers to hemodialysis and the reasons for such transfers. If a peritoneal dialysis-related infection resulted in loss of the catheter and thus transfer to hemodialysis, the specific type of infection was recorded as the reason for the transfer. Transfer to hemodialysis was considered permanent when a CAPO patient switched to hemodialysis with no intention of returning to peritoneal dialysis. Transfer was considered temporary when a CAPO patient transferred to hemodialysis but later returned to CAPO. A category of no transfer was designated when a CAPO patient never received hemodialysis after CAPO was initiated. Some of the patients in the no transfer group underwent hemodialysis before the initiation of CAPO. Statistical analysis was performed using X2 and analysis of variance. Results are expressed as mean values + SD. Infection rates (catheter infection and peritonitis) were calculated as total number of infections for all patients in a group divided by total time of continuous peritoneal dialysis. Because infection rates decreased with increasing CAPO experience, the rates were calculated separately for patients receiving peritoneal dialysis treatment for < 1 year, 1 to 2 years, and > 2 years so that valid comparisons of the no transfer, permanent transfer, and temporary transfer groups could be made.

RESULTS

The characteristics of those peritoneal dialysis patients who permanently, temporarily, and never transferred to hemodialysis are shown in Table 1. Table 1.

Patients, n Months of CAPD/patient * Age, yr Black patients, n (0/0) Diabetic patients, n (0/0) Female patients, n (%)

Twenty-seven percent (68/247) of the patients permanently transferred to hemodialysis after a mean time of 15 + 14 months (range, five days to 6 years) of continuous peritoneal dialysis. Another 29% (72/247) transferred at least once temporarily to hemodialysis. The proportion of black patients who permanently transferred was higher than the proportion of nonblack patients who transferred (18/35 v 50/212; P < 0.01). Nineteen percent of insulin-dependent diabetics permanently transferred to hemodialysis, and 32 % of nondiabetics permanently transferred (16/83 v 52/164; 0.1 > P > 0.05). The ratio of men to women was the same in all three groups. Catheter infections were the leading cause for both permanent and temporary transfer to hemodialysis, as shown in Table 2. Catheter infections (with or without peritonitis) accounted for 56% (65/115) of the temporary and 37 % (25/68) of the permanent transfers from CAPD. Peritonitis was the second most frequent cause for permanent and temporary transfer to hemodialysis. Infection rates and numbers of patients in each group, analyzed by time on CAPD, are shown in Fig 1. Fifty-six percent (38/68) of the patients who permanently transferred to hemodialysis did so in the first year of CAPD, 22 % (15/68) in the second year, and another 22 % after > 2 years of CAPD. In contrast, the percentage of patients requiring temporary hemodialysis increased with increasing time of CAPD. Catheter infection rates were higher in patients who transferred to hemodialysis than in those who never transferred when analyzed by time spent on CAPO (Fig 1). For patients undergoing peritoneal dialysis for < 1 year, catheter infection rates were three times higher in the patients who permanently or temporarily transferred to hemodialysis than in

Patient Characteristics

Permanent Transfer to Hemodialysis

Temporary Transfer to Hemodialysis

68

72

15 + 10 50 + 16

35 + 23 46 + 15

18 (26)t

7 (10)

No Transfer to Hemodialysis

107 17 + 15 49 + 16 10 (9)

16 (23)

25 (35)

42 (39)

28 (41)

28 (39)

40 (37)

* Analysis of variance, P < 0.01. tx 2 = 10.4; P < 0.01; ratio of black to white patients for those permanently transferring to hemodialysis

v all others.

TRANSFER OF CAPO PATIENTS TO HEMODIALYSIS .

Table 2.

367

Reasons for Transfer to Hemodialysis

Reason

Infection Catheter alone Peritonitis alone Both Patient preference Leak or no drain Noncompliance Inadequate clearance and/or ultrafiltration Other

Temporary Transfers

Permanent Transfers

37 27 28 0 20 0

15 -13 10 9 4 7

0 3

6 4

115*

Total

patients who permanently transferred to hemodialysis in our CAPD population. Peritonitis uncomplicated by catheter infection accounted for an additional 19 % of permanent transfers. Others have documented the importance of peritonitis as a factor resulting in the transfer of patients to hemodialysis but generally have not listed catheter infections as a separate cause of transfer. 3-6 Nissenson et al found that of their CAPD patients who permanently transferred to another dialysis method, 56 % did so because of peritonitis or catheter infections or both 2 ; these results were similar to ours. Two multicenter studies, one American 3 and one British, 4 have shown that patients who transfer to hemodialysis have very high peritonitis rates. Our results confirm the hi~h peritonitis rate in patients who leave CAPD, but additionally demonstrate that catheter infections are also elevated in these patients compared with those of patients continuing to receive CAPD. There are several explanations for catheter infection, rather than peritonitis, being the major cause for permanent transfer of CAPD patients to hemodialysis at our center. We identified the specific CAPD infections (peritonitis, catheter infection, or both), resulting in removal of patients from CAPD, rather than grouping all such infections together. If peritonitis episodes associated with catheter infections had been considered only as "peritonitis," then peritonitis would have appeared to be the more important cause of catheter loss and permanent transfer to hemodialysis. We were aggressive about removing peritoneal catheters for persistent catheter inf~ctions because we felt this prevented the development of peri-

68

*Seventy-two patients transferred temporarily to hemodialysis for a total of 115 temporary transfers.

those who did not (3.0 v 3.3 v O. 9 episode/year). Catheter infection rates decreased with increasing time of CAPD. However, even those patients who transferred to hemodialysis after 2 years of peritoneal dialysis had a catheter infection rate almost double that of those who did not transfer to hemodialysis (1.1 v 1.1 v 0.6 episode/year; permanent v temporary v no transfer, respectively). Likewise, peritonitis rates were higher in patients who did than in those who did not transfer to hemodialysis, but the differences were not as striking as for catheter infection rates. DISCUSSION

We found catheter infections to be even more important than peritonitis as a reason for the permanent transfer of CAPD patients to hemodialysis. Catheter infections, either directly or because of associated peritonitis, accounted for 37% of the A.

CATHETER INFECTIONS

B.

PERITONITIS

I~:/}(rl PERMANENT _

a:

4.0

TEMPORARY

~ NO TRANSFER

«w

>-

a: 3.0 w

a.. en

z

02.0 Fig 1. Infection rates, as episodes per year, in patients grouped by time of CAPO and by whether they transferred (permanently or temporarily) to hemodialysis. The number of patients for each period is shown below each column.

t5w u.

z 1.0

O.OL...-.,j~

No. PATIENTS 38 15 48 <1 YEAR

15

9 38

1 -2 YEARS

15 48 21

< 1 YEAR TIME ON CAPD

>2 YEARS

1 -2 YEARS

>2 YEARS

368

toni tis. One fifth of the peritonitis episodes at our center are associated temporally and by microbiologic results with catheter infections. 8,9 We have a high catheter infection rate (1.1 episode/ year) in our population and a relatively low peritonitis rate (0.9 episode/year). In contrast, the national peritonitis rate is 1.4 episode/year and catheter infection rate is 0.5 episode/year.1o An additional reason we found catheter infections to be an important cause of transfer from CAPO is the way we classified reasons for such transfer. Categories used by the CAPO National Registry as reasons for transfer to hemodialysis included "medical reasons other than peritonitis" and patient choice. 1 Both categories could include patients who transferred because of catheter infections. The patients at our center who transferred permanently to hemodialysis because their catheters were removed for persistent catheter infection could have returned to CAPO but chose not to do so. We did not attribute this CAPO failure to patient preference or to other medical reasons but to catheter infection. Catheter infection rates decreased with increasing time of peritoneal dialysis (Fig lA). We found that the catheter infection rate was strikingly elevated in patients who permanently transferred during their first year of CAPO, being three times the rate of the patients continuing to receive CAPD (3 episode/year v 0.9 episode/year). The catheter infection rate was 1.1 episode/year in those patients who discontinued peritoneal dialysis after 2 years compared with 0.6 episode/year in those patients continuing CAPO. These results suggest that recurrent catheter infections, especially early in the patients' CAPD experience, discouraged the patients, and many then elected to discontinue CAPO permanently. The patients continuing CAPD were the ones with the lower infection rates, which accounts for the decrease in infection rates with increasing time of peritoneal dialysis. Twenty-seven percent (68/247) of our patients (mean time of CAPO, 22 months) transferred permanently to hemodialysis. These results are similar to those of Nissenson et ai, who reported that 26 % of their CAPD patients (mean time of CAPO, 12 months) transferred to hemodialysis. 2 Most of our permanent transfers occurred during the first year of CAPO (38/68, or 56% of all permanent transfers). Thirty-eight percent of patients receiving CAPD for < 1 year (38/101) transferred per-

PIRAINO, BERNARDINI, AND SORKIN

manently to hemodialysis as opposed to only 18 % (15/84) of those patients receiving CAPD for> 2 years who then transferred permanently from CAPD. Steinberg et al also found that the percent of patients transferred permanently to hemodialysis decreased with increasing time of CAPO. 1 Very little data have been published on the reasons and percentage of patients temporarily (as opposed to permanently) transferring from CAPO to hemodialysis. Gokal et al found that 233 temporary hemodialysis transfers were needed in 610 CAPO patients, on dialysis for a mean of 1 year, representing 0.4 transfers per patient per year of CAPO.4 We found a similarly high percentage of patients requiring temporary hemodialysis. Of 247 patients receiving CAPD for a mean ot 22 months, 29% required temporary hemodialysis. This represents 0.25 temporary transfers per patient per year. An additional 27 % of our patients transferred permanently to hemodialysis, leaving only 43 % who had never required hemodialysis by the time the study ended. Gokal et al noted that dialysis programs must plan for back-up in-center hemodialysis for their CAPO patients in view of the large percentage of patients who will require this. 4 Our study confirms the importance of maintaining the availability of in-center hemodialysis space for CAPO patients. We found catheter infecti~ns to be the leading reason for temporary hemodialysis. In contrast, Gokal et al found peritonitis to be the CUlprit in the majority of instances. 4 This disparity may be partially because we counted catheter infections resulting in peritonitis (24 % of transfers) as a category separate from peritonitis without catheter infection (23 % of transfers). Catheter infections alone accounted for another 32 % of transfers. Mechanical problems related to the catheter accounted for the remainder of the temporary transfers in this center and elsewhere. 4 When we looked for other factors effecting transfer from CAPD, we found that black patients were more likely to transfer permanently than all others. We have no explanation for this. Black patients do not appear to be at increased risk of catheter infection.7 However, our findings agree with those of the CAPO National Registry, which reported similar results. 3 Further studies will be required to determine why black patients are more likely to transfer permanently to hemodialysis than other patients.

TRANSFER OF CAPO PATIENTS TO HEMODIALYSIS ·

369

We examined the effect of diabetes (regardless of cause of chronic renal failure) on permanent transfer to hemodialysis. Nineteen percent of the diabetics (all but two receiving intraperitoneal insulin) transferred permanently off CAPD, in contrast to 32 % of all other patients (16/83 v 52/164; P < 0.1). This trend for diabetics to be more likely than others to continue CAPD may be in part because of the good glucose and blood pressure control such patients have while receiving CAPD. 11-13 Other studies have not shown diabetics to be at either increased or decreased risk of transfer. 3, 14 However, one study 3 did not include patients who developed diabetes mellitus subsequent to initiation of CAPD as diabetics, as we did. Twelve percent (10/83) of our diabetics did not have diabetes as the cause of their end-stage renal disease. We previously have not found insulin-dependent diabetics to be at increased or decreased risk for catheter infections. 7 Further study is

needed to determine if diabetics are less likely to transfer from CAPD than other patients. This study emphasizes the importance of catheter infections as a cause of transfer of CAPD patients to hemodialysis. Further work needs to be directed toward the prevention of catheter infections. Recent reviews examine possible ways to prevent catheter infections. 15 Placement of the catheter, care of the exit site, and choice of catheter all are of potential importance in minimizing infections. New catheters that are more resistant to catheter infections are currently being evaluated. 17 ACKNOWLEDGMENT We are indebted to the CAPD nurses of the University Health Center of the Pittsburgh Dialysis Unit and the Veterans Administration Medical Center for their excellent patient care. Our thanks to Lucy Perrotta for her assistance in the preparation of the manuscript, Dr August Turano for his invaluable assistance in the computers and the file manager system, and Dr lean Holley for reviewing the manuscript.

REFERENCES 1. Steinberg SM, Cutler Sl, Nolph KD, et al: A comprehensive report on the experience of patients on continuous ambulatory peritoneal dialysis for the treatment of end-stage renal disease. Am 1 Kidney Dis 4:233-241, 1984 2. Nissenson AR, Gentile DE, Soderblom RE, et al: Morbidity and mortality of continuous ambulatory peritoneal dialysis. Regional experience and long-term prospects. Am 1 Kidney Dis 7:229-234, 1986 3. Nolph KD, Cutler Sl, Steinberg SM, et al: Factors associated with morbidity and mortality among patients on CAPD. Trans Am Soc Artif Intern Organs 33:57-65, 1987 4. Gokal R, King 1, Bogle S, et al: Outcome in patients on continuous ambulatory peritoneal dialysis and hemodialysis: 4year analysis of a prospective multicentre study. Lancet 2:1105-1109, 1987 5. Tranaeus A, Heimburger 0, Lindholm B, et al: Six years' experience of CAPD at one center; A survey of major findings. Perit Dial Intern 8:31-41,1988 6. Khanna R, Wu G, Vas S, et al: Mortality and Morbidity on CAPD. ASAIO S 4: 197-204, 1983 7. Piraino B, Bernardini 1, Sorkin M: The influence of peritoneal catheter exit -site infections on peritonitis, tunnel infections, and catheter loss in patients on continuous ambulatory peritoneal dialysis. Am 1 Kidney Dis 8:436-440, 1986 8. Piraino B, Bernardini 1, Sorkin MI: A 5 year study of the microbiologic results of exit site infections and peritonitis in CAPD. Am 1 Kidney Dis 10:201-286, 1987

9. Bernardini 1, Piraino B, Sorkin MI: Analyses of CAPDrelated Pseudomonas aeruginosa infections. Am 1 Med 83: 829-832, 1987 10. Lindblad AS, Novak lW, Nolph KD, et al: The 1987 USA National CAPD Registry report. Trans Am Soc Artif Intern Organs 34: 150-156, 1988 11. Flynn CT, Shadur CA: A comparison of continuous ambulatory peritoneal dialysis in diabetic and nondiabetic patients. Am 1 Kidney Dis 1: 15-23, 1981 12. Madden MA, Zimmerman Sw, Simpson DP: Continuous ambulatory peritoneal dialysis in diabetes mellitus. Am 1 Nephrol 2:133-139, 1982 13. Amair P, Khanna R, Leibel B, et al: Continuous ambulatory peritoneal dialysis in diabetics with end-stage renal disease. N Engl 1 Med 306:625-630, 1982 14. Rubin 1, Kirchner K, Ray R, et al: Demographic factors associated with dialysis technique failure among patients undergoing continuous ambulatory peritoneal dialysis. Arch Intern Med 145: 1041-1044, 1985 15. Copley lB: Prevention of peritoneal dialysis catheter-related infections. Am 1 Kidney Dis 10:401-407, 1987 16. Oreopoulos DG, Baird-Helfrich G, Khanna R, et al: Peritoneal catheters and exit-site practices: Current recommendations. Perit Dial Bull 7:130-138, 1987 17. Trooskin SZ, Donetz AP, Baxter 1, et al: Infection-resistant continuous peritoneal dialysis catheters. Nephron 46:263267, 1987