NDT Advance Access originally published online on October 11, 2006
Nephrology Dialysis Transplantation 2007 22(2):545-551; doi:10.1093/ndt/gfl563
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Adherence to peritoneal dialysis training schedule
Department of Medicine & Therapeutics, Prince of Wales Hospital, The Chinese University of Hong Kong, Shatin, Hong Kong, SAR, China
Correspondence and offprint requests to: Dr K. M. Chow, Department of Medicine & Therapeutics, Prince of Wales Hospital, The Chinese University of Hong Kong, Shatin, Hong Kong, SAR, China. Email: Chow_Kai_Ming{at}alumni.cuhk.net
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and methodsResultsDiscussionAcknowledgementReferences |
|---|
Background. Shortening behaviour during peritoneal dialysis training can be easily measured, and likened to the skipping behaviour in haemodialysis subjects, although its effect on peritoneal dialysis outcomes is now well understood. We studied the clinical impact of failing to adhere to a peritoneal dialysis training programme among incident dialysis patients.
Methods. This study included 159 consecutive inception peritoneal dialysis patients in a single centre from September 1999 through November 2002. We evaluated the effects of behavioural compliance quantified by the per cent time arriving late for scheduled peritoneal dialysis training. The patients were categorized by whether they arrived late in >20% of their peritoneal dialysis training sessions.
Results. Of the 159 incident peritoneal dialysis patients (mean age 57 ± 13 years) who attended peritoneal dialysis training, 70 subjects (44%) arrived late in >20% of the sessions. They were younger by 5 years than patients who arrived late 
20%. Mean peritonitis-free time for subjects who arrived late for training in >20% the of sessions was 30.9 months, as compared with 41.8 months in subjects with 20% late attendance behaviour (log rank test, P = 0.038). Multivariable Cox proportional hazards analysis showed that late attendance behaviour and baseline serum albumin were the only independent risk factors for the time to a first peritonitis after adjustment for diabetes mellitus and relevant coexisting medical factors. Late arrival in >20% of the peritoneal dialysis training sessions was associated with >50% increased likelihood of subsequent peritonitis, with an adjusted risk ratio of 1.56 (95% confidence interval, 1.02–2.39; P = 0.04).
Conclusion. These findings show that the behavioural measure of late attendance for peritoneal dialysis training has a crucial role in predicting peritonitis. It may therefore represent a practical strategy for identifying poor adherence or predicting medical outcomes.
Keywords: adherence; behavioural; compliance; dialysis training; peritoneal dialysis; peritonitis
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionAcknowledgementReferences |
|---|
Because end-stage renal disease therapy requires a multidisciplinary approach of care, there are numerous ways in which dialysis patients can have non-adherence [1,2]. The ability of nephrologists to recognize and assess adherence is poor. Peritoneal dialysis patients, as opposed to haemodialysis patients, are spared the twice-weekly or thrice-weekly visits, and obviously cannot be assessed by the pattern of skipping dialysis treatment session or inter-dialytic weight gain. Use of computerized printouts of dialysis solution deliveries, home supply inventories by home visits, and questionnaires during follow-up visits have been proposed but all of these correlates are confounded by factors other than adherence.
Short of a good standard to measure adherence in peritoneal dialysis patients [2], indicators of non-adherence to treatment regimen are nevertheless considered a useful resource for nephrologists to help identify patients who are most in need of interventions. In particular, the personality trait of conscientiousness, among a five-factor personality trait (a hierarchical organization of personality traits in terms of five basic dimensions including extraversion, agreeableness, conscientiousness, neuroticism and openness to experience), is the only dimension which correlates with adherence among dialysis patients [3]. To minimize the confounding of adherence measures with biological phenomena, increasing attention has centred on the use of behaviourally based measures to denote adherence. Such a measure should be easily interpretable, unambiguous in meaning and under direct patient volition. While the schedule of in-centre haemodialysis treatments offers a unique opportunity to examine non-adherence behaviours such as arriving late or leaving early [1,4], the degree of skipping treatment is relatively difficult to assess in the setting of home-based peritoneal dialysis therapy. To circumvent the shortcomings, we operationalized the definition of attendance by means of detailed training attendance record for incident peritoneal dialysis patients. Using their behaviour in following the peritoneal dialysis training schedules as a surrogate, we examine the clinical impact of behavioural factors on peritoneal dialysis outcomes. Our hypothesis is driven by the correlation of human behaviour and compliance demonstrated previously [3,5]. First, we examined the hypothesis that the behaviour of arriving late for dialysis training would be associated with an increased likelihood of adverse outcomes such as peritonitis, increased serum phosphate levels and hospitalization. Second, we hypothesized that such behaviour would be predictive of death risk. Better understanding of such behavioural factors will define future targets for multidisciplinary management of end-stage renal disease.
|
Subjects and methods |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionAcknowledgementReferences |
|---|
This is a prospective observational study, in which the study population consisted of all incident end-stage renal disease patients who began peritoneal dialysis at our unit between September 1999 and November 2002. Patients were followed up until death or 31 September 2005, at which point data were censored.
Availability of detailed attendance records on each patient allowed us to further categorize them according to whether they adhered to the training schedule starting from 9 a.m. All patients received a written and verbal instruction to attend training before 9 a.m. The training lasted for 5 h each day on average. Patients attending the scheduled training later than 9 a.m. as instructed were classified as late arrival (irrespective of the amount of time lag). The number of late arrivals was divided by the total days of peritoneal dialysis training. Patients were characterized by whether they arrived late in >20% of their scheduled peritoneal dialysis training sessions. This cutoff value was selected based on the distribution of late attendance in this population and a clinical rule of thumb that those who arrive late >20% would shorten training more than once out of a 5-day training course on average. For instance, if a patient attended peritoneal dialysis training later than 9 a.m. on two occasions out of 6 days, he or she was classified as late arrival in >20%. In this way, the proportion of patients who were classified by the cutoff 20% late attendance came close to the median according to the late attendance ranking. Because our definition of late attendance by the cutoff value of 20% is arbitrary, we further performed analyses by categorizing the subjects with late attendance in 20, 20–40 and >40% of the training sessions.
Peritoneal dialysis was performed by means of disconnect systems [Baxter, SpA (Deerfield, IL) and Fresenius Medical Care (Deutschland GmbH) with lactate-buffered glucose-containing dialysate solutions]. Prophylactic antimicrobial cefazolin was administered prior to Tenckhoff catheter placement in all cases. Dialysis was initiated after a break-in period of 4 weeks after surgical placement of the catheter. The standard continuous ambulatory peritoneal dialysis (CAPD) training programme lasted for 3–14 days. Topical mupirocin, with or without oral rifampin, was administered in the case of Staphylococcus aureus nasal carrier with S. aureus exit-site infection or peritonitis. After the initial home visit during the first month after training, follow-up home visits were not routinely arranged.
Data were obtained from the patient records and renal replacement therapy registry electronic databases. The medical data collected consist of demographic details, marital status, employment status, information about the primary renal disease, range of comorbidities (coronary artery disease, peripheral artery disease, cerebrovascular disease, chronic obstructive airway disease, hepatitis B surface antigen status and diabetes mellitus), and anthropometric data at dialysis initiation. Baseline comorbidity was also assessed at the start of dialysis by the Charlson Comorbidity Index validated for the peritoneal dialysis population [6]. Baseline serum albumin level was measured by means of the bromcresol purple method 1 month after initiation of dialysis. To account for the travel time to the dialysis unit, we obtained proxy measures of journey time by documenting each person's residential address. Patients were then classified by their geographic living location—whether they lived within the catchment area of our hospital cluster spanning 300 km2 (population 1 300 000).
In terms of dialysis outcomes, we measured the chance of developing dialysis-related peritonitis, phosphate control and the risk of hospitalization, all-cause mortality and cardiovascular mortality. Times to first peritonitis episode were examined using standard survival methodology. Serum phosphate levels were analysed by using the mean values for 3 months at 1 year after the commencement of the dialysis. Hospitalizations were recorded, and expressed as adjusted days in hospital per patient-year after renal replacement therapy. Cardiovascular mortality included death associated with a definite myocardial ischaemic event, heart failure, cerebrovascular accident, arrhythmia or peripheral vascular accident, all of which were defined according to standard clinical criteria and sudden death, which was defined as unexpected natural death within 1 h from the symptom onset and without any prior condition that would appear fatal [7].
Statistical analysis was performed by SPSS for Windows software version 11.5 (SPSS Inc., Chicago, IL). All data were expressed as mean ± SD for normally distributed data and median or range for skewed data. Statistical comparisons were performed using unpaired Student's t-test; comparisons of percentages between groups were made with the chi square test or Fisher's exact test as appropriate. Kaplan–Meier survival plots were used to display hypothesized relationships, and stepwise multivariate Cox regression proportional hazards model was used to predict hazard of developing dialysis-related peritonitis. Analyses were censored at death, transplantation and transferal to haemodialysis or other modalities of renal replacement therapy. The dependent variable was the time to a first dialysis-related peritonitis. The Cox models were adjusted for covariates that were believed to be potential confounders for the time to the first peritonitis. On the basis of our work [8, 9] and that of others [10, 11, 12], variables used for modelling were patient age, baseline serum albumin level, diabetic status, Charlson Comorbidity Index and employment status. As for the outcomes of cardiovascular and all-cause mortality, survival curves were built with the Kaplan–Meier method and according to the peritoneal dialysis training attendance behaviour. All probabilities were two-tailed and the level of significance was set at 0.05.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionAcknowledgementReferences |
|---|
During the study period, a total of 170 incident Chinese end-stage renal disease patients were enrolled in the peritoneal dialysis programme. After exclusion of subjects who had no training attendance data from a computerized medical record, 159 patients were included in the analyses. This study population (mean age 57 ± 13 years, range 20–87 years) was studied over a total observation period of 519 patient years. The median follow-up period was 38 months. The subjects were 48% male; 48% had diabetes and 14% required helpers for dialysis exchanges. Only 4% of the subjects received automated peritoneal dialysis. During the observation period, 62 patients died and 28 of them underwent kidney transplant.
The average peritoneal dialysis training lasted for 5.1 ± 1.5 days. Seventy (44%) incident peritoneal dialysis patients arrived late in >20% of the training sessions. On average, they were late in 33 ± 20% of the sessions. As quantified by the minutes of late arrival throughout the training period, the first group (late arrival >20%) presented late by a median of 43 min in total, significantly more than the group with late arrival 20% (P < 0.0001). Subjects who arrived late for training in >20% of the sessions were more likely to be younger (59 ± 15 vs 54 ± 15 years, P = 0.046), although we found little association between late arrival and the employment status. Characteristics of two groups of subjects are described in Table 1. There was no gender predilection (P = 0.87). In addition, no difference in prevalence of diabetes, preexisting cardiovascular disease, cerebrovascular disease and other comorbid conditions, as measured by the baseline Charlson Comorbidity Index scores, was found. At the time of peritoneal dialysis training, the majority (99%) of patients who arrived late in >20% of the training sessions were living within the catchment area of our hospital, as compared with 92% of the cases who arrived late in 20% (P = 0.08).
|
The overall peritonitis rates were 35.9 patient-month/episode and 21.1 patient-month/episode for subjects with
20% and >20% late attendance, respectively. To evaluate the peritoneal dialysis outcomes, risk of developing the first peritonitis was estimated with respect to the training attendance behaviour. During the period of observation, 92 patients (58%) had developed dialysis-related peritonitis. Sixty-two episodes (67%) were due to gram-positive organisms. The Kaplan–Meier method was used to generate unadjusted estimates of peritonitis-free survival. Time to an initial episode of peritonitis varied significantly with respect to the adherence to peritoneal dialysis training schedule (Figure 1). Life table analysis of mean peritonitis-free time for subjects who arrived late for training in >20% of the sessions was 30.9 months, which was worse than subjects with 20% late attendance behaviour, who had mean peritonitis-free time of 41.8 months (log rank test, P = 0.038). In particular, this effect was not observed in gram-negative peritonitis (P = 0.17) but largely confined to gram-positive peritonitis episodes (log rank test, P = 0.0013; Figure 2). Technique of survival did not differ between the two groups (log rank test, P = 0.58). We re-analysed gram-positive peritonitis risk among three subsets of patients with late attendance in 20% (n = 89), 20–40% (n = 56), >40% (n = 14) of the training sessions. Compared with patients with the lowest late attendance, those with an increasing degree of late attendance behaviour carried a progressively heightened risk of gram-positive peritonitis (log rank test, P = 0.0035; Figure 3). Risk of gram-positive peritonitis was also analysed with reference to the categories of patients who arrived late by more and less than 20 min per session on average (Figure 4). Although the trend did not reach statistical significance, subjects who arrived late by >20 min per training session tended to develop gram-positive peritonitis earlier.
|
|
|
|
Multivariable Cox proportional hazard model was used to determine the independent effects of late attendance behaviour in predicting peritonitis and included terms for age, baseline serum albumin level, diabetic mellitus, Charlson Comorbidity Index and employment status. From this Cox regression analysis (Table 2), the two independent predictors of peritonitis were late attendance behaviour and baseline serum albumin level. Late arrival in >20% of the peritoneal dialysis training sessions significantly increased the likelihood of subsequent dialysis-related peritonitis, with an adjusted risk ratio of 1.56 (95% confidence interval, 1.02–2.39; P = 0.04). Baseline serum albumin level at 1 month after initiation of dialysis also independently predicted peritonitis after dialysis therapy; the relative risk of peritonitis was 2.00 (95% confidence interval, 1.27–3.16; P = 0.003) for every 10 g/l decrease in serum albumin.
|
Time-averaged serum phosphate concentration for 3 months at 1 year after commencing dialysis did not differ between two groups of patients, 1.63 ±0.57 mmol/l for patients with
20% late arrival vs 1.63 ± 0.48 mmol/l for those with >20% late arrival (P = 1.00).
The median number of days of hospitalization per year was similar between the two groups after starting maintenance peritoneal dialysis: 7.1 days (interquartile range, 1.2–16.3 days) vs 6.0 days (interquartile range, 0.4–19.7 days) among the patients with >20% late arrival and 20% late arrival respectively (P = 0.97).
Finally, no association was found between the late attendance behaviour for peritoneal dialysis and cumulative probability for all-cause mortality (log rank test, P = 0.99) or cardiovascular mortality (log rank test, P = 0.81). Table 3 shows the causes of death in our study cohort. During the observation period, we observed a slightly higher chance for patients with >20% late attendance to receive renal transplants (23% vs 13%, P = 0.15).
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionAcknowledgementReferences |
|---|
In this large cohort of incident end-stage renal disease patients requiring peritoneal dialysis, we have applied detailed analysis in an attempt to evaluate the implication of peritoneal dialysis training behaviour on clinical outcomes and risk of peritonitis. Our findings support the hypothesis that the individual's attitude or behaviour is predictive of peritonitis risk among patients on peritoneal dialysis. We showed that patients with >20% late arrival during peritoneal dialysis training period were associated with a statistically significant 56% increase in risk for peritonitis after controlling for age, diabetes mellitus, baseline serum albumin levels and other potential risk factors for dialysis-related peritonitis. In particular, the risk association holds true for gram-positive peritonitis but not gram-negative peritonitis (which are often subject to non-behavioural influences). Our findings are noteworthy because the risk of gram-positive peritonitis increased progressively with a higher per cent of late attendance behaviour.
Although our data cannot directly address the reason for the association of training attendance behaviour with peritonitis, several explanations are possible. We sought but did not find a link between employment or housewife status with late arrival. Likewise, residential address outside the hospital catchment area, known to increase the travel time, was not observed to predict late attendance for peritoneal dialysis training. One can speculate that the self-discipline qualities implicit in the peritoneal dialysis training attendance behaviour may influence a patient's compliance with dialysis exchange procedures. Conversely, lack of conscientiousness can derail both the commitment and the ability to follow through with peritoneal dialysis treatment. Although non-adherence may be one important mediator of the link between late attendance behaviour and poor health outcomes in peritoneal dialysis, it has been difficult to prove or refute a missing link of depression or psychosocial factors. Since depression among peritoneal dialysis patients is a recognized risk factor for peritonitis [13,14], it seems plausible that high peritonitis rate among patients who had late attendance for training programme is related to underlying depression. By extension, patient behaviour of late arrival for dialysis training could be considered a composite or summated effect of adherence or motivation, psychological problems (particularly depression), patient's lack of belief in benefit of peritoneal dialysis procedures, and provider–patient relationship.
We showed no correlation of training attendance behaviour with the serum phosphate levels 1 year after peritoneal dialysis. It appears that significant residual renal function among incident peritoneal dialysis patients could have a greater influence on phosphate control than behavioural characteristics. These findings further emphasize the construct validity of behavioural measures, which might perform better than conventional laboratory indices to measure adherence or compliance. Furthermore, the late attendance behaviour might predict the adherence to following the instructed steps of sterile dialysis exchange technique in the first place, but not the compliance with performing dialysis which should better be verified with the use of home visit supply inventories [15].
It should be kept in mind that subjects with >20% late attendance tended to have a higher chance of receiving kidney transplants in this study. This could have explained, in part, the observation that late attendance behaviour did not predict survival or hospitalization risk.
Potential limitations of our study should be noted. Our record did not allow us to differentiate between excused and unexcused late attendance for training, although there is no reason to suspect the effect of physical illness or social support. Of those with >20% late arrival behaviour, they were neither more likely to require helpers nor have higher comorbid disease burden. As in other retrospective studies, although this analysis establishes the association between peritoneal dialysis training schedule attendance behaviour and the subsequent risk of peritonitis, it cannot ascertain causative relationships. There is another caveat to this study. Although one nurse was responsible for the training of each peritoneal dialysis subject, the patient: caseload ratio was relatively great (patient to nurse ratio 50 : 1) within our unit. Important questions remain concerning the validity of extrapolating our results to dialysis centres with higher staffing levels. Finally, as our multivariate models depend on the variables that are available, certain potential confounders may not be included in the analysis (for example, exit-site infection status, depression score, educational level). Analysis of the role of these variables may present a new and exciting opportunity for future research.
In summary, the present study demonstrated a novel behavioural factor predictive of dialysis-related peritonitis independent of known risk factors. We believe that with better understanding of patient behaviour and peritoneal dialysis adherence, it would be of clinical impact on further refinement of patient care. Although most physicians are confident that they can predict the adherence of their patients with therapy, the predictions are no more accurate than can be obtained by the toss of a coin [16]. This study might provide the impetus for future intervention using simple rules to identify and target the at-risk patient group shortly after initiating dialysis. An important consideration is the potential use when applying our data to clinical practice. Clearly, our findings may prompt many nephrologists to consider using our approach to estimate the adherence of patients with the proper performance of peritoneal dialysis exchanges at home. Because late arrival in >20% of the training sessions appeared to predict the future risk of peritonitis, a strong argument can be made that further home visits or technique review should be considered after completion of training in this at-risk patient group. It seems clear that further study is necessary to determine whether adherent behaviour can be taught and whether behavioural modification strategies reduce peritonitis risk and improve outcomes in the peritoneal dialysis population.
|
Acknowledgement |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionAcknowledgementReferences |
|---|
This study was supported in part by the CUHK research accounts 6900972 and 6900570.
Conflict of interest statement. None declared.
|
References |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionAcknowledgementReferences |
|---|
- Kaveh K and Kimmel PL. (2001) Compliance in hemodialysis patients: multidimensional measures in search of a gold standard. Am J Kidney Dis 37:244–266.[Web of Science][Medline]
- Leggat JE Jr. (2005) Adherence with dialysis: a focus on mortality risk. Semin Dial 18:137–141.[CrossRef][Web of Science][Medline]
- Christensen AJ and Smith TW. (1995) Personality and patient adherence: correlates of the five-factor model in renal dialysis. J Behav Med 18:305–313.[CrossRef][Web of Science][Medline]
- Unruh ML, Evans IV, Fink NE, Powe NR, Meyer KB. (2005) The Choices for Healthy Outcomes in Caring for End-Stage Renal Disease (CHOICE) study. Skipped treatments, markers of nutritional nonadherence, and survival among incident hemodialysis patients. Am J Kidney Dis 46:1107–1116.[CrossRef][Web of Science][Medline]
- Wright N and Tanner MS. (2002) Medical students’ compliance with simple administrative tasks and success in final examinations: retrospective cohort study. Br Med J 324:1554–1555.
[Free Full Text] - Beddhu S, Zeidel ML, Saul M, et al. (2002) The effects of comorbid conditions on the outcomes of patients undergoing peritoneal dialysis. Am J Med 112:696–701.[CrossRef][Web of Science][Medline]
- Engelstein ED and Zipes DP. (1998) Sudden cardiac death. In Alexander RW, Schlant RC, Fuster V (Eds.). The Heart, Arteries and Veins.(McGraw-Hill, New York) pp. 1081–1111.
- Chow KM, Szeto CC, Leung CB, Kwan BC, Law MC, Li PK. (2005) A risk analysis of continuous ambulatory peritoneal dialysis-related peritonitis. Perit Dial Int 25:374–379.
[Abstract/Free Full Text] - Chow KM, Szeto CC, Leung CB, Law MC, Li PK. (2005) Impact of social factors on patients on peritoneal dialysis. Nephrol Dial Transplant 20:2504–2510.
[Abstract/Free Full Text] - Farias MG, Soucie JM, McClellan W, Mitch WE. (1994) Race and the risk of peritonitis: an analysis of factors associated with the initial episode. Kidney Int 46:1392–1396.[Web of Science][Medline]
- Oxton LL, Zimmerman SW, Roecker EB, Wakeen M. (1993) Risk factors for peritoneal dialysis-related infections. Perit Dial Int 14:137–144.
- Wang Q, Bernardini J, Piraino B, Fried L. (2003) Albumin at the start of peritoneal dialysis predicts the development of peritonitis. Am J Kidney Dis 41:664–669.[CrossRef][Web of Science][Medline]
- Juergensen PH, Wuerth DB, Juergensen DM, et al. (1997) Psychosocial factors and clinical outcome on CAPD. Adv Perit Dial 13:121–124.[Medline]
- Troidle L, Watnick S, Wuerth DB, Gorban-Brennan N, Kliger AS, Finkelstein FO. (2003) Depression and its association with peritonitis in long-term peritoneal dialysis patients. Am J Kidney Dis 42:350–354.[CrossRef][Web of Science][Medline]
- Bernardini J, Nagy M, Piraino B. (2000) Pattern of noncompliance with dialysis exchanges in peritoneal dialysis patients. Am J Kidney Dis 35:1104–1110.[Web of Science][Medline]
- Stephenson BJ, Rowe BH, Haynes RB, Macharia WM, Leon G. (1993) Is this patient taking the treatment as prescribed? JAMA 269:2779–2781.
[Abstract/Free Full Text]
Accepted in revised form: 4. 9.06
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  K. M. Chow, C. C. Szeto, M. C. Law, J. S. Fun Fung, and P. Kam-Tao Li Influence of Peritoneal Dialysis Training Nurses' Experience on Peritonitis Rates Clin. J. Am. Soc. Nephrol., July 1, 2007; 2(4): 647 - 652. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||