Connecting devices in CAPD and their impact on peritonitis.A Shetty, DG Oreopoulos
Toronto Hospital (Western Division), Ontario, Canada., Canada
Keywords: Clinical Trials, Equipment Contamination, Human, Peritoneal Dialysis, adverse effects,instrumentation,standards,Peritonitis, etiology,prevention &control,Risk Factors,
In 1976 when CAPD was first described, two one litre glass bottles were used with a long disposable transfer set. Each exchange with this system involved a large number of steps and hence the peritonitis rate was high (one episode every three months). Since then, CAPD has advanced to the point where it has become an acceptable mode of treatment for patients with end - stage renal disease. However, in spite of all these advances peritonitis remains a major cause of morbidity and of drop - out from therapy in patients on CAPD. An early breakthrough in the control of peritonitis was the introduction of the Toronto Western Hospital technique for CAPD, on Sept 26, 1977. This involved a wearable 2 litre plastic bag connected to the PD catheter through a permanent tube. This technique decreased the average peritonitis rate from one episode every three to one every 12 patient months.
In 1992, 65000 patients on CAPD all over the world testify to the success of CAPD following the application of the Toronto Western Hospital (TWH) technique. The original TWH technique, even though, it had fewer steps/exchanges had several disadvantages such as (1) the mandatory first step after the connection is made (the point of possible contamination) is the flow into the peritoneal cavity of the fresh dialysate and (2) the necessity of carrying the empty bag and the connecting tube at all times.
In 1980, Buoncristiani of Perugia, Italy introduced the Y or the Perugia system, which markedly reduced peritonitis rates, although it took a long time for the rest of the world to be convinced about the efficacy of this simple modification of the original technique. Today, all nephrologists agree with the principles of the Perugia System. The key to the success of the Y - system is the flush before fill" concept and the use of a disinfectant maintained in the line during the entire dwell time which is washed out as the first step of each new bag exchange. It has not been established that it is essential to use the disinfectant and a variety of systems now in use that do not require disinfectant, still have low peritontis rates. The Y set is a simple Y - shaped tube. It has three branches, one of which is connected to the catheter through a titanium Luer - lock adaptor; of the remaining two branches, one is connected to the fresh dialysate bag and the other to a drainage bag or an open container that collects the drained dialysate.
The "O - set", is a modification of the Y- set, wherein it has a Y - configuration during use, but assumes "O" shape between exchanges when the two branches of the Y are connected. In the Ultraset R, another modification & the Y-set, the effluent bag is incorporated into the connecting tubings. Before the exchange, the Y-set is connected to the new bag and a small amount of fluid is flushed over: then, after drainage of the peritoneal cavity, the new solution is infused. This disposable set uses no antiseptic agent because the whole system is discarded. A further modification of the Y set is the twin bag system in which the dialysate bag and the drainage bag come already attached to the Y shaped tubings. This removes the spiking step thus all that is required is to connect the tubing to the catheter and then disconnect it at the same place. Even though Buzzato et al introduced the twin bag system as early as 1980, for years the system was not successful chiefly because the only commercial model available at that time employed respectively a needle and a rubber cap as connectors at the bag and at the catheter side. This made the system less effective in preventing peritonitis because the flush was not complete due to the cul-de-sac formed between the needle and rubber cap. Now that the needle and rubber cap, have been replaced by Luer-lock connectors, peritonitis rates have declined considerably.
These disconnect systems have several advantages: a) they make it unnecessary to wear the bag, b) when volume reduction is required, they give the patient the ability to change the fill volume without ordering bags of different sizes, and, c) they may reduce shear trauma to the exit site and thus decrease exit-site infection. The possible disadvantages of the disconnect systems are: 1) increased cost per exchange for disposable supplies; 2) risk of accidental sodium hypochlorite infusion, whenever this is used as a disinfectant, risking chemical damage to the peritoneum. The advantages of twin bags are: 1) single Luer-lock connection, 2) single stage and aseptic manipulation, 3) short training period, 4) good patient acceptance, 5) low incidence of exogenous peritonitis, 6) this modification can be used even by patients with poor manual dexterity or those with visual or mental impairment because it does not require the precision that spiking requires. Its main disadvantage is increased cost.
Although bacteria gain entry to the peritoneal cavity through different routes, the intraluminal route is believed to be particularly important . In the standard connection system, once the spike connection is made, any bacteria that have been introduced into the system, flow directly into the peritoneal cavity with the dialysate. With the Y - set, the first step after spike bag connection, is to flush out the potentially contaminated dialysate in the down- flow tube of the dialysate bag with about 100 ml of fresh dialysate. The next step is to flush the catheter and the Y-set by draining out the spent dialysate from the peritoneum. In the Y system with disinfectant, these two steps also serve to flush away the disinfectant. It has not yet been demonstrated whether the second step has any additional benefit, e.g. by flushing away any bacteria from the catheter and the Y-set. In Ultraset R and twin bag systems, which involve connecting the tubing to the catheter proximal to the "Y" during each exchange, this second flush likely flushes away any residual contamination.
In vitro studies of flushing effect
The efficacy of flushing for removing bacteria depends upon the adherent properties of micro-organisms, the quantity of dialysate used for flushing, and the length of time between Inoculation and flushing. Studies in which before bag connection, a Y-set branch was contaminated with 200 colony - forming units (CFU) of Pseudomonas aeruginosa, showed that flushing with 5 ml of fresh dialysate removes up to 50% of the bacteria and flushing with 110 mi removed all of the bacteria. Verger and Luzar showed that flushing with 100 ml of fluid immediately after inoculation removed 100% of Staph epidermidis but only 60% of Staph aureus and 30% of Ps aeruginosa. The efficacy of flushing done 10 hours after contamination remained at 100% for Staph epidermidis but was reduced to zero for Staph aureus and Pseudomonas aeruginosa ganisms. Immediate flushing is analogous to flushing out the contamination while connecting a disposable system. Delayed flushing is analogous to flushing the contamination from reusable systems like O-set, and removing the contamination that occurs at the time of capping the catheter at the end of the dialysis exchange. Verger et al also showed that Staphylococcus epidermidis did not migrate along the catheter lumen but 60% of Pseudomonas did - an observation that suggests that we need a disinfectant to kill these migrating organisms, which are introduced at the end of dialysis. It remains to be seen whether the Betadine in the betacap system that covers the titanium adaptor at the end of the catheter is sufficient or whether we need In-line disinfectant.
There is as yet no agreement on whether a disinfectant is necessary. The arguments in favour of a disinfectant are: 1) Flushing is not 100% effective In washing out all the organisms, 2) delayed flushing does not (0%) wash away Staph aureus, Pseudomonas aeruginosa and possibly many organisms other than Staph epidermidis. Thus flushing does not protect against contamination by these organisms if it occurs after completion of each exchange - that Is at the time of capping. Some of these organisms like Pseudomonas can migrate along the catheter lumen into the peritoneum, 3) no other group that does not use disinifectant in the system have reported the very low peritonitis rate described by Buoncristiani and Maiorca both groups that use disinfectant.
Also it is not clear whether this high success rate is due principally to the disinfectant. The arguments against the use of a disinfectant are: 1) Risk of its accidental introduction into the peritoneal cavity with potential long-term damage to peritoneal membrane function, 2) immediate painful chemical peritonitis that occurs at least once in 25% of patients, 3) increased complexity of the procedure that makes it difficult to learn 4) risk of chronic anxiety concerning accidental introduction of the disinfectant and 5) absence of strong convincing in vivo evidence of its usefulness.
Dasgupta ‘found that after contamination, flushing reduced both bacterial growth and the development of bacterial biofilm; the addition of the disinfectant bleach conferred an added benefit, eradicating both bacteria and biofilm. In the original Y system, the antiseptic was sodium hypochlorite (Amuchin) used undiluted or at 50% concentration. In an in vitro study, 5 x 109 CFU of Pseudomonas aeruginosa were sterilized in less than 30 seconds by 5 ml of a 4.5% concentration, and in 5 minutes by a 0.7 percent concentration of Amuchin. On rare occasions disinfectant may be introduced accidentally into the peritoneal cavity. In a large multi-centre study of the "O" set system Villano reported that this occurred once in every 4380 bag exchanges. Maiorca et al reported a frequency of one episode every 7500 bag exchanges, i.e. one episode per 62.4 patient months.
The Canadian CAPD Clinical Trials Group, found accidental infusion of sodium hypochlorite once in 2500 exchanges; that is, it occurred at least once in 25% of patients. Swartz et al reported accidental sodium hypochlorite infusion once in 6 patients and twice in 2 patients over the 542 patient months (no. of patients 35). i.e. once per 8000 CAPD exchanges. It is of concern that in 2 patients, accidental infusion contributed to the discontinuation of the O-set: in one due to persistent sterile peritonitis and identification of a large ovarian cyst following accidental infusion, and in the second at the patient's request. Even though the risk of accidental infusion is small, it may work to lower the acceptability of CAPD because this is an iatrogenic complication and the pain associated with it usually is severe enough to be remembered for months or years.
Maicirca et al found no reduction in peritoneal creatinine clearance or in ultrafiltration in those patients who had accidental infusion of the disinfectant into the peritoneal space. In spite of these results in a small number of patients, most nephrologists, including ourselves are not convinced that the chemical burn caused by the disinfectant has no significant long-term effects on the peritoneum. For these reasons any system that has a disinfectant within the lumen has not gained wide acceptance. Since we do not have convincing evidence that a disinfectant in the line is better than no disinfectant, we have concluded that it is not necessary.
The literature,,,,,,,,,,,,,,,,,, makes it obvious that 'flush before fill" is better than standard spiking. The peritonitis rate in patients using the Y-set or Y-systern was one episode per 10.4-69.4 patient-months, with a median value of 1 episode every 27.65 months (mean 1 episode per 32.02 months). In contrast the peritonitis rate in patients using the standard spiking is one episode per 7.57 to 13 months, with a median of 1 episode every 10 months (mean 1 episode per 10.4 months, From the in vitro results of flushing, we expect almost total eradication of Staph epidermidis peritonitis and a marked reduction in the peritonitis caused by Staph aureus, Pseudomonas aeruginosa and possibly other organisms. After a multi-centre controlled study, the Canadian CAPD Clinical Trials group reported that, in the Y system group, peritonitis due to epidermidis was greatly reduced (2/21 peritonitis episodes for the Y system group and 21/47 for standard system group). Peritonitis due to other organisms also was reduced but to a lesser extent. Cantaluppi et al also found marked reduction in staph epidermidis peritonitis. In most of the other studies, peritonitis rate due to all organisms was reduced by the same extent. Even though Y set uniformly fared better than the standard spike system, there was a wide variation in the peritonitis rates in the different studies.
The various explanations for this variation are as follows:
1) In his initial report Buoncristiani included only those who were culture positive. Thus culture-negative peritonitis, which the other groups include was not included in Buoncristiani's study. However, even if these patients are included, Buoncristianj still had very low peritonitis rates.
2) Confer effects - An interesting finding in the Italian CAPD Cooperative Study Group and the Canadian study is the variability of results from centre to centre. Most of the other studies that reported very good results ware prospective studies from highly experienced teams, In contrast to this, the Italian Cooperative retrospective CAPD study included data from all centres including those with little experience with peritoneal dialysis. This may explain a peritonitis rate of one episode per 17.3 months compared to much better results in other studies. In the Canadian multi-centre study, two centres had no episodes of peritonitis in 60 patient months.
3) Disinfectant - An analysis of the studies that report peritonitis rates greater than one episode per 36 months have one important factor in common, namely the use of in-line disinfectant. No study of a disposable connecting system using the flush-before-fill concept without any in-line disinfectant has reported such a high success rate. Evidence of the in vitro efficacy of the disinfectant suggests that thq presence of disinfectant improves the outcome in these patients. Also it is possible that the mere presence of the Y set is responsible for lower rates of peritonitis.
4) All these studies excluded peritonitis-related to exit - site infection and/or tunnel infection. Peritonitis due to tunnel infection is easy to diagnose. Peritonitis dire to exit - site infection is a poorly defined entity because peritonitis caused by skin organisms is very common even in the absence of exit-site infection. Because exit-site infection also is common, it is possible that the exit-site infection and the peritonitis often are caused by the same organism. Also it is possible that more peritonitis episodes have been attributed to exit-site infection in these studies that report a high success rate.
No centres in North America or UK have been able to reproduce the impressively low peritonitis rates reported from Italy, even after using in) line disinfectant.
There is reason to believe that the latest twin bag system will be better than UltrasetR because everything in Twin bag is similar to the UltrasetR, except that, in the former, the dialysate bag is preattached to the tubing (viz there is no possibility of contamination during the spiking stage). The peritonitis rates reported for UltrasetR are one episode per 10.4 to 27.79 months (14, 22-24): those for the twin-bag system are one episode per 25.9 - 34.3 months,,. Dratwa et al reported a peritonitis rate of one episode per 29.4 months in 18 patients using the Twin BagR system for 235 patient months, this can be compared with one episode per 10.4 months in 15 patients who used UltrasetR for 198 patient months. UK peritonitis experience was one episode per 34.3 months over 890 patient months on CAPD using Twin BagR system, compared with one episode per 26.9 months in those who used the Y set for 1211 patient months. Bazzato et al reported a peritonitis rate of one episode per 26.6 months in 16 patients who received long term CAPD (i.e. for more than 4 years). The latter study probably suffers from a selection bias because the study included only those who were on long term CAPD. i.e. only those who otherwise were doing well.
There has been no formal controlled trial between UltrasetR and the Twin BagsR system but North American centres using these two systems appear to have comparable results and hence eventually the twin bag system will become the system of choice.
Sterilizing and assist devices represented another attempt to reduce the contamination that occurred at the time of connecting the bag. These devices serve a double purpose: they assist those with poor manual dexterity and/or poor eyesight in performing the exchanges; they act as sterilizing devices to decontaminate the connecting site. The utraviolet-exchange device (UV-XD) employs the germicidal effectiveness of UV light and the sterile connecting device (SCD) uses a heated blade to cut and weld (and sterilize) the tubes. Other sterilizing devices include the Hong Kong connection and flame lock system. Of all these devices, the only one that is in common use is the W-M UV irradiation in this device is produced by a beam at 254 wave length for 30 seconds in the connector box. In living organisms UV light disrupts nucleic acids and proteins; the most effective wave lengths are between 253 and 265 mm31. However, in a multi-centre randomized Study. UV-XD did not reduce the peritonitis rate. The mean time to peritonitis did not change in a group after transfer to an UV-XD. A few other smaller studies, have shown some benefit of UV over standard spiking. Few studies have compared UV device with V set. During 974 patient months on O set and 1010 patient months on UV device (UV-XD) Bailie et al found an odds ratio indicating increased risk of developing peritonitis in patients using UV-XD. In two different groups of patients, Tjandra et al found a peritonitis rate of one episode per 25.9 months during 467 patient-months on the twin-bag system and one episode per 18.6 months during 817 patient months on the UV device. In contrast, Poseno et al found higher peritonitis rate in those who used the O set compared to those who used UV-XD, however, the duration of follow-up was short and the number of patients who used the O set was very small. Stermayr et al tried to determine whether the main cause for the reduction in peritonitis was the mechanical effect of the UV- box or the M light. In this study, patients agreed to use the UV device without UV light for 6 months and then to start the UV device with UV light. There was a significant reduction in the peritonitis rates from 1 episode per 10 months to 1 episode every 17 months after transfer to the UV light. The authors concluded that UV box improves the peritonitis rates but that these improve further after using the UV light, Peritonitis rate in our patients using UV flash is also quite high (unpublished data). It is difficult to explain why we cannot get a reasonable success rate with UV device. Perhaps those patients using UV flash are high risk patients (which is true because only those with poor vision and poor manual dexterity get to use the UV flash) or the UV device does not efficiently sterilize the connecting site. It is quite probable that many patients do not use the UV flash properly because of their poor physical condition. It seems unlikely that, if it is used properly, the UV device would fail to sterilize the connecting site. Frequently on inquiry, we find patients who admit that they did not use the UV flash for spiking. The intermediate group of patients with slightly impaired vision (but not blind) and slightly poor manual dexterity may be able to use the twin bag system, which may help reduce the peritonitis rate in such patients. In those with grossly impaired vision UV flash is still the device of choice.
The "flush-before-fill" concept should be applied in all the possible connection devices. The necessity of in-line disinfectant has not yet been established. The twin-bag system probably will replace all other systems in the near future. Blind patients and those with poor manual dexterity may still have to continue with assist devices, such as UV flash, in spite of high peritonitis rates.
Our sincere thanks to Cristy Espino and Margarida Silva for their help in preparing this manuscript and to Dr. JO Godden for his editorial review of the final manuscript.