Original Article. Christian Verger 1, Mirela Duman 1, Pierre-Yves Durand 2, Ghislaine Veniez 1, Emmanuel Fabre 1 and Jean-Philippe Ryckelynck 3

Transcription

1 Nephrol Dial Transplant (2007) 22: doi: /ndt/gfl760 Advance Access publication 31 January 2007 Original Article Influence of autonomy and type of home assistance on the prevention of peritonitis in assisted automated peritoneal dialysis patients. An analysis of data from the French Language Peritoneal Dialysis Registry Christian Verger 1, Mirela Duman 1, Pierre-Yves Durand 2, Ghislaine Veniez 1, Emmanuel Fabre 1 and Jean-Philippe Ryckelynck 3 1 Centre hospitalier Rene Dubos, Pontoise, 2 ALTIR, Vandoeuvre and 3 CHU Cle menceau, Caen, France Abstract Background. In France, 48% of home-based peritoneal dialysis (PD) patients require assistance to perform their exchange and manage their treatment. A total of 7% are aided by their family, and 41% by a private nurse. Of all the continuous ambulatory peritoneal dialysis (CAPD) patients, 61.7%, and among automated peritoneal dialysis (APD) patients 23%, are assisted at home for their bag exchanges and connections. Assisted APD patients (AAPD) are more comorbid and elderly so that a home helper is not always available: this explains why most helpers at home are private visiting nurses paid by the National Social Security. In addition to the home helper (nurse or family), 58% of centres make regular additional home visits to check the respect of procedures previously taught during the initial training of the nurse or the family helper. The aim of this study was to evaluate whether the type of home assistance received by dependent patients had an influence on peritonitis rates, and if home visits done by nurses of training centres may improve results. Methods. Peritonitis rates and the probability of being peritonitis free were analysed for 1624 new APD patients recorded in the French PD Registry (RDPLF) between 2000 and 2004, and followed-up until early Results. Nurse-assisted APD patients had a peritonitis rate of one episode every 36 months, and familyassisted patients one episode every 45 months; using Poisson analysis this trend was not significant (P ¼ 0.11). However, the probability of being peritonitis free was significantly higher for familyassisted (69.8% at 2 year) compared with home Correspondence and offprint requests to: Docteur Christian Verger, RDPLF Service de Dialyse, Centre Hospitalier Rene Dubos, 6 Avenue de l Ile de France Pontoise (France). c.verger@wanadoo.fr nurse-aided persons (54.4%) after adjustment for age, diabetes and the Charlson comorbidity index. This difference disappeared when nurses from the training centre regularly visited PD patients at their home in the presence of their helper, whichever type of assistance they received. In addition, when the nurses from the training centres visited private nurse-assisted patients, the probability of being peritonitis free was significantly improved in comparison with those persons who did not receive home visits, from 33.9% to 50.8% at 3 years (P ¼ 0.028). Conclusions. APD patients assisted at home by a private nurse have a higher risk of developing peritonitis than family-assisted patients, unless additional regular home visits are organized by the original training centre. Therefore, we recommend that home visits be regularly made for dependent PD patients to optimize the quality of care provided by the helper. Keywords: assisted peritoneal dialysis; autonomy; elderly patients; home dialysis; peritoneal dialysis; peritonitis; registry Introduction A high percentage of peritoneal dialysis (PD) patients in France are elderly, frequently suffer from numerous comorbidities, and are unable to perform their treatment by themselves. We recently reported that 56% of French PD patients need some degree of assistance at home, so that PD exchanges can be performed correctly [1]. Seven percent are aided by their family, 41% by private nurses, and 3% received assistance of not-defined type of helper. Most of these at home assisted patients are on CAPD (61.7% of CAPD patients in the French PD Registry in 2005), whereas only 23% of automated peritoneal dialysis (APD) patients are assisted. ß The Author [2007]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org

2 Type of home assistance and peritonitis risk on peritoneal dialysis 1219 This home assistance is of two types: it is either provided by a family member, or by a private nurse who is paid by the French National Health Service [2]. The French Language Peritoneal Dialysis Registry (RDPLF) has recorded data on about 82% of French PD patients. The information available in its database includes, among other, the type of dependency, the presence or not of diabetes, the Charlson comorbidity index and peritonitis rates. The aim of this study was, based on the aforementioned RDPLF data, to compare the risk of developing peritonitis in PD patients who were either assisted at home by a family member or aided by a private nurse, respectively. As there is an increasing trend in other European countries to organize nursing assistance for dependent APD patients in their own home, rather than leaving them aided by a family member, we considered that this study might provide some useful findings for such ongoing and future projects. We restricted our study to patients treated by automated peritoneal dialysis. In the following text, we call AAPD (assisted automated peritoneal dialysis) the non-autonomous patients treated at home on APD for whom the connections of the lines and preparation of the machine are made by a family member or a private home nurse. Subjects and methods Study population The RDPLF gathers data on PD patients from metropolitan France, the French departments and territories, and also from Algeria, Argentina, Belgium, Switzerland, Tunisia and Uruguay. However, for the purposes of the present study, only PD patients undergoing treatment in metropolitan France were selected, as patients treated in the French departments and territories or abroad are exposed to different climatic conditions and socio-economic environments which could conceivably influence the results, including the description of the patient population as well as the quality and type of home assistance. Only APD patients are being examined in this study. The French PD Registry has recorded PD patient data since However, for this investigation, in order to present an analysis of the most recent results, we limited our selection to incident adult PD patients who started PD treatment between 1 January 2000 and 1 December 2004 and were followed-up until 1 January In all, 1624 APD patients from 112 different centres were included. Details on the study populations are given in Table 1. Peritonitis rates Peritonitis rates were calculated by adding together all the periods of treatment on the PD systems investigated and dividing it by the number of episodes of peritonitis that occurred while on these systems. The probability of being peritonitis free was calculated using actuarial curves; for this calculation, patients were followed-up as of the first episode of peritonitis, or up to the time they stopped PD or were lost to follow-up. The date of the first episode of peritonitis was not censored, whereas patients without any episode of peritonitis were censored when they stopped PD or were lost to follow-up, or at the end of the study period. For patients on APD, the machines and systems used are the same, whatever the degree of patient dependency and type of home assistance. All patients use luer lock connectors. The evaluation of the probability of being peritonitis free was adjusted for age, presence of diabetes, and the Charlson comorbidity index using the multivariate Cox model. In AAPD patients, the difference in peritonitis rates was analysed using Poisson analysis. Results are presented in Table 1. The groups of AAPD patients assisted by family Table 1. Main characteristics of incident French metropolitan APD patients who started a PD treatment between January 2000 and December 2004 Variable All APD Non-assisted Home nurseassisted Home family-assisted patients patients patients n ¼ 1624 n ¼ 1265 (78%) n ¼ 232 (14%) n ¼ 127 (8%) Age (years): mean SD Male: n (%) 1036 (63.8%) 835 (66%) 128 (55.2%) 73 (57.5%) Diabetics: n (%) 299 (18.4%) 191 (15.1%) 68 (29.3%) 40 (31.5%) Charlson index a 4 (2 15) 4 (2 14) 7 (2 15) 6 (2 11) Median (range) Centre size b <20 patients 637 (39.2%) 528 (82.9%) 70 (11.0%) 39 (6.1%) patients 370 (22.8%) 265 (71.6%) 76 (20.5%) 29 (7.9%) >30 patients 617 (37.9%) 472 (76.5%) 86 (13.9%) 59 (9.6%) Peritonitis rate (One episode 1/34 1/33 1/36 1/45 per n patient-months) Probability of being peritonitis free (95% CI) at 24 months 58.1% ( %) 59.3% ( %) 54.4% ( %) 69.8% ( %) at 36 months 45.3% ( %) 45.6% ( %) 39.8% ( %) 52.1% ( %) Home visit c Yes 967 (59.5%) 760 (60.1%) 130 (56.0%) 77 (60.6%) No 657 (40.5%) 505 (39.9%) 102 (44.0%) 50 (39.4%) a Charlson index is evaluated at initiation of treatment. b Centre size is defined by the number of incident patients who started PD between 2000 and c Training Centre make home visits or not.

3 1220 C. Verger et al. or nurses are also described in Table 1. We used Student t- test for the quantitative variables and chi-square test for the qualitative variables to compare these groups. As a bias might be attributed to a centre effect, mainly its experience (which may be partly deducted from the number of patients trained), the calculations were repeated using a centre variable expressing the size of the centres based on the number of new patients included by each centre during the study period. The different groups are presented in Table 2. To evaluate the influence of the centre size variable, we used the Cox model. Additional questionnaire An additional questionnaire was sent to all the centres, in which we asked if the centre had set up the organization of home visits made by a nurse from the centre itself in the presence of either the family or that of the home nurse, etc. All 112 centres included in the study replied to this questionnaire: 58% of the centres included in the survey make home visits while 42% do not. When home visits are made by the nurses of the training centre, the rhythm may vary from one centre to another, i.e. from only once at initiation to several times per year. In the questionnaire, we asked to reply yes only if centres were doing home visits on a regular basis, that is to say at least once a year. Usually it is done every 3 or 6 months. During these visits, management of the treatment, technique of connection and aseptic procedures are checked. This is done as well in autonomous patients as with the family helper or the home-nurse helper in non-autonomous patients. All home nurses must have a previous training in the dialysis centre before they can care for PD patients at home. It must be underlined that for these nurses, PD patients are only part of their daily work load as they maintain their conventional activity, that is to say caring for non-renal patients who can have various other diseases, needing for example chemiotherapy, medical injections, skin injury care, patient washing, etc. Most of time the initial training of a family helper takes 1 2 weeks, whereas the training of home private nurses is two to five 1 or 2 h sessions. The probability of being peritonitis free was then analysed again by introducing into the Cox model the answer to the question do nurses from the centre make regular visits to the patient s home? for all categories of patients (non-assisted, assisted by a family member, or assisted by a private home nurse). The analysis was carried out on a Macintosh Power PC using JMP statistical software. Results Table 1 shows that 22% of patients included needed an assistance at their home to perform their PD exchanges and manage their treatment: 8% were helped by their family and 14% were aided by a private nurse. Overall peritonitis rates Overall peritonitis rates on APD amounted to one episode every 34 patient-months (cumulative duration of treatment: months). Peritonitis rates for patients on APD Peritonitis rates for patients on APD amounted to one episode per 33 patient-months for autonomous patients (with a total cumulative duration of treatment amounting to patient-months), one episode every 45 patient-months for patients assisted at home by a family member (with a total cumulative duration of treatment amounting to 2565 patient-months), one episode per 36 patient-months for patients assisted at home by a private nurse (with a total cumulative duration of treatment amounting to 2844 patient-months). Albeit differences in peritonitis rates represent a trend, it was not significant when using Poisson analysis (P ¼ 0.11) Probability of being peritonitis-free for patients on APD Influence of dependency, comorbidities and centre size. Figure 1 provides an outline regarding the probability of being peritonitis free for both non-assisted and assisted APD patients. It should be noted that significantly better results were observed for patients assisted by a family member compared with those assisted by a private nurse. The variable, centre size, had no influence on the results (Table 2). The variable, home visit or not by nurses from the training centre, was introduced into the model and the probability of being peritonitis free was analysed again including this additional parameter. Table 2. APD: Probability of being peritonitis free depending on the centre size (number of incident patients who started PD between January 2000 and December 2004) Centre size Number of patients Probability of being peritonitis free (95% CI) P-value at 24 months at 36 months < % ( %) 41.5% ( %) to % ( %) 44.5% ( %) > % ( %) 47.9% ( %)

4 Type of home assistance and peritonitis risk on peritoneal dialysis 1221 Fig. 1. Analysis of data from the French PD Registry regarding the probability of being peritonitis free for new automated peritoneal dialysis (APD) patients who started their treatment between 2000 and 2004, and who were followed-up until early Percentages regarding autonomous patients, persons assisted by a family member and those assisted by a private nurse are shown. Table 3. Main characteristics of patients whether the training centre nurses makes home visits or not Variable Home visits No visits P-value Age (years): mean SD Male: n (%) 119 (57.5%) 82 (53.9%) Diabetics: n (%) 138 (66.7%) 113 (74.3%) Helper Home nurse 130 (62.8%) 102 (67.1%) Family assisted 77 (37.2%) 50 (32.9%) for family-assisted patients, i.e % at 2 years, vs % for those aided by a home nurse had a probability of being peritonitis free (P ¼ 0.015). When patients were assisted by a family member, no difference was observed regarding the results for centres, whether or not they were sending one of their nurses to make a home visit. When patients were assisted by a private nurse, the results regarding the probability of being peritonitis free were better in centres that sent one of their nurses to make a home visit (Table 4). Influence of home visits on the results. The findings regarding patients assisted by a family member or by a private nurse have been summarized in Table 1. It should be noted that there were significantly more aged PD patients in the nurse-assisted group of patients. Table 3 compares characteristics of patients where the training centres make home visits (207 AAPD patients) vs those who do not (152 AAPD patients), for the variables age, sex and frequency of diabetes, in patients assisted by family and nurse. When home visits were also made by one of the nurses from the training centre, no significant difference was found between PD patients assisted by a family member or by a private nurse: the probability of being peritonitis-free amounted to % at 2 years for the patients aided by a family member, and to % for those helped by a home nurse, respectively (P ¼ 0.58). When home visits were not made by a nurse from the training centre, a significant difference in the results was observed between those PD patients assisted by a family member and those aided by a private nurse. In this case, the results were significantly better Discussion As previously mentioned, France is not the only European country to organize private nurses to treat home-based PD patients, and reports on previous experiences have already been published in the literature [3,4], with satisfactory results. However, compared with other countries that have adopted this approach, France has used the latter extensively due to the selection of older patients [5 6]. As patient age increases, so do the number of comorbid conditions as well as the degree of dependency. In addition, the spouses of elderly PD patients are often dependent themselves, or may have already died. This partly explains, with both CAPD and APD included, why around 41% of all PD patients in France are treated with the help of a private nurse. About 7% are aided by a member of the family. The remaining percentage of patients is autonomous, and does not require assistance. Recently, Lobbedez et al. [7] demonstrated that patients assisted by a nurse had only a 50% probability of being peritonitis free at 12 months.

5 1222 C. Verger et al. Table 4. Probability of being peritonitis free at 24 and 36 months for AAPD patients when assisted by a private nurse, whether the training centre makes home visits or not Type of Centre Number of patients Probability of being peritonitis free (95% CI) P-value at 24 months at 36 months With home visits % ( %) 50.8% ( %) Without home visits % ( %) 33.9% ( %) However, in their experience, as in most other centres in France [1], the method of choice for assisted patients was CAPD. In France, APD is mostly prescribed in younger patients who want to maintain their autonomy during the day and keep on working full time. This explains why the percentage of AAPD patients is lower in the RDPLF compared with assisted CAPD patients. However, the increasing lack of nurses who accept visits to the patient s home 3 or 4 times per day might change in the near future: hence, with AAPD the nurse has to come only twice a day, and even sometimes only once in the evening to mount the machine and connect the patient, whereas the patient may clamp the line at the end of the session in the morning and disconnect by himself without risk; in addition, the trend in other European countries to increase AAPD may change the profile of the patients in the future, so the French experience in this field, obtained through the results of the RDPLF, may be helpful. The initial purpose of this study was to determine whether AAPD patients helped by a non-professional, i.e. a member of the family, had a higher risk of developing peritonitis. The Poisson analysis did not demonstrate a significant difference in peritonitis rates, albeit a trend to a higher rate was observed when dialysis procedures are performed at home by the private nurse: it must be underlined however, that Poisson analysis has limitations and it has been shown, using data from a multicentre trial, that the risk of peritonitis is best described in terms of survival curves analysing probability of being peritonitis free rather than the mean peritonitis rates and Poisson analysis [8]. Thus, for the purpose of this study, the probability of being peritonitis free appeared to us as a more reliable way to analyse the influence of home assistance in dependent patients. In fact, better results were obtained when the PD patient was assisted by a family member rather than by a trained professional, i.e. a private nurse: these findings were rather surprising. For patients on APD, the successive analyses showed that the results were equivalent for centres that sent one of their nurses to visit home-based patients on a regular basis, both when the latter were assisted by a family member or by a private nurse. It was also found that these visits significantly improved the results regarding the private nurse s performance, but not those concerning the family member. Different explanations may be put forward. First of all, a family member only takes care of a single patient towards whom there is a high degree of personal involvement. He or she is given initial training, and this is regularly re-evaluated during visits to the out-patient clinic, as usually this family member accompanies the PD patient at all consultations. In addition, in most cases it is nearly always the same member of the family who cares for the patient, so he or she gains a certain degree of experience, and profits from continuous training. On the contrary, private nurses cannot concentrate on just one PD patient. They have to take care of a number of different patients, and a PD patient is frequently an exception within a total number of 20 to 40 patients with other pathologies who have to be followed per day. Moreover, the order in which the private nurses treat their patients may have some importance. It would be preferable, if possible, for the private nurse to first visit the PD patient, as there is a risk that she might unknowingly transfer a bacterial infection from a patient with another pathology, whom she has just treated. Unlike the family member, after their initial training, private nurses do not return to the out-patient clinic with the PD patient, so retraining is not necessarily an easy matter. In addition, due to the lack of time and the number of patients they have to treat, private nurses are not that keen on having a retraining course, or a re-evaluation of how they perform PD exchange. Last but not least, most of these private nurses usually work in groups of three to five: some of them change from time to time, and it could quite well happen that a substitute receives training without the team informing the centre in question; the private nurses may therefore teach their new colleague some variations on the original technique which was explained to them during their initial training by the nursing team from the training centre. Then again, their equipment and the way in which it is used may not meet hospital standards: they may tend to adapt what they have been taught in the hospital to the home environment: however, as this may be done without the control of a specialized PD nurse, there is a danger that these adaptations may increase the risk of contamination, and therefore of infection. Whereas 80% of the training centres have their nurses who make home visits in the United States [9], only 58% of the French centres were found to do so in our survey. However, it should also be noted that the influence of home visits in reducing the risk of

6 Type of home assistance and peritonitis risk on peritoneal dialysis 1223 peritonitis has been seldom evaluated [10]. The results of setting up this system of monitoring, retraining and providing professional encouragement, should be further examined; hence, when home visits are performed by one of the nurses from the training centre, she brings her experience in situ to the private home nurse, helps her to adapt the PD exchange procedure to the home environment, and at the same time ensures that the basic principles of peritonitis prevention are adhered to. This collaboration between private home nurses and hospital nurses who visit them on a regular basis is generally well accepted. The analysis of data from the registry presented in this study therefore confirms that the results regarding private nurses and family members are similar in situations where the former can benefit from the periodical re-evaluation of their practice at home, with in situ support received from their hospital colleagues. It should be underlined that even when private nurses do not benefit from a regular home visit by a nurse from the centre, the peritonitis rates are still far below the minimum level determined by the ISPD recommendations which is less than one episode every 18 months [11]. Therefore, if a home visit on the part of a training centre nurse is not possible, it can nevertheless be stated that the level of security for PD patients cared for by home nurses remains very satisfactory, and that the relevant results meet the criteria laid down in best practice guidelines. Conclusion In France, the possibility of having private nurses to care for dependent home-based PD patients allows nearly 40% of all PD patients to be treated. Whereas, we observed similar trends in CAPD patients (personal unpublished RDPLF report), we focused this study on automated peritoneal dialysis patients who represent a more homogenous population in terms of PD systems used. The overall results, in terms of peritonitis prevention, meet the international standards. To optimize the results obtained by private nurses we recommend that, whenever possible, regular home visits should also be made by the hospital nurses in order to help the private nurses maintain and even further their knowledge and practice of PD exchanges. When such a continuous evaluation program is set up, similar results are then obtained both for private nurses and for family members. The results are even better than those for non-assisted PD patients, and allow non-autonomous patients to remain home based and yet with a high level of security. Acknowledgements. The Registre de Dialyse Pe ritone ale de Langue Franc aise (RDPLF) si supported by an annual grants of the following companies (French affiliates): AMGEN, BAXTER, GAMBRO, GENZYME, FRESENIUS, ROCHE, SHIRE, THERABEL-PHARMA. We would also like to express our thanks to Mrs Anne-Marilyn Schreier-Audoire for her assistance with language correction. We warmly thank all nurses and physicians who participate to the RDPLF and have made this study possible. The names of the 120 participating centres are available on the internet site Conflict of interest statement. None declared. References 1. Verger C, Ryckelynck JP, Duman M et al. French peritoneal dialysis registry (RDPLF): outline and main results. Kidney Int 2006; 70: S12 S20 2. Durand PY, Verger C. Commentary: the current state of PD in France. Perit Dial Int 2006; 26: Povlsen JV, Ivarsen P. Assisted automated peritoneal dialysis (AAPD) for the functionally dependent and elderly patient. Perit Dial Int 2005; 25 [Suppl 3]: S Sunder S, Taskapan H, Jojoa J et al. Chronic peritoneal dialysis in the tenth decade of life. Int Urol Nephrol 2004; 36: Vrtovsnik F, Porcher R, Michel C et al. Survival of elderly patients on peritoneal dialysis: retrospective study of 292 patients, from 1982 to Perit Dial Int 2002; 22: Issad B, Benevent D, Allouache M et al. 213 elderly uremic patients over 75 years of age treated with long-term peritoneal dialysis: a French multicenter study. Perit Dial Int 1996; 16 [Suppl 1]: S Lobbedez T, Moldovan R, Lecam M, Hurault de Ligny B, El Haggan W, Ryckelynck JP. Assisted peritoneal dialysis. Experience in a French renal department. Perit Dial Bull 2006; 26: Vonesh E. Estimating rates of recurrent peritonitis for patients on CAPD. Perit Dial Bull 1985; 5: Bernadini J, Dacko C. A survey of home visits at peritoneal dialyis centers in the United States. Perit Dial Int 1998; 18: Castro MJ, Celadilla O, Munoz I et al. Home training experience in peritoneal dialysis patients. EDTNA ERCA J 2002; 28: Piraino B, Bailie GR, Bernardini J et al. Peritonteal Dialysisrelated infections. Recommendations: 2005 update. Perit Dial Int 2005; 25: Received for publication: Accepted in revised form: