Abdominal pain is a common complication of peritoneal dialysis (PD), with a multifactorial etiology1,23. For example, it has been reported that abdominal pain occurs in 78% of patients during an episode of peritonitis4. However, pain experienced during intraperitoneal infusion of the dialysis solution also occurs in the non-peritonitis state, and is distinguished by the abdominal pain appearing within the first minute or so after starting infusion of the PD solution, but which generally diminishes during the dwell5. Infusion pain is usually observed in new patients commencing dialysis and is often transient in nature, spontaneously disappearing over time3,6. However, it is established that infusion pain can remain a troublesome complication in PD patients and in the most extreme cases can result in the discontinuation of PD7.
Infusion pain is generally believed to be due to the acidity (pH 5.2 to 5.5) of conventional lactate buffered dialysis solutions. Such solutions not only contain high concentrations of lactate, but also glucose and are acidic to minimize caramelization as well as the chemical transformation of glucose to 5-hydroxymethylfurfural and its acidic metabolites7,8. Other factors that may contribute to infusion pain include the position of the catheter, the temperature of the dialysis solution2, and the presence of high glucose concentrations in the PD fluid, as it is not uncommon for patients to predominantly experience pain during the infusion of hypertonic solutions7.
A number of actions can be employed to alleviate infusion pain. By far the most common treatment is the neutralization of the PD solution by the manual injection of sodium bicarbonate into the solution immediately prior to infusion. Sodium bicarbonate is preferred to other bases, such as hydroxide, or sodium carbonate because of its safety, buffering action and gradual titration curve9. A number of reports have demonstrated the effectiveness of such manual injection in alleviating infusion pain8,10,11. However, the need to inject into every bag prior to infusion adds to the therapy burden for the patient and also increases the overall treatment cost. The greatest concern, however, is related to the possibility that peritonitis rates are greatly increased in patients who manually inject sodium bicarbonate due to external contamination of the solution. For example, in a study by Henderson, Couper and Lumsden7, the peritonitis rate in 10 patients injecting bicarbonate for the treatment of inflow pain was 1 episode/4 patient months. Following cessation of bicarbonate treatment, the peritonitis rate decreased to 1 episode/9 patient months. Other possible actions that may be employed to treat inflow pain include slowing the infusion rate, incomplete drainage (including the use of tidal therapy in automated peritoneal dialysis), manual injection of local anesthetics into the solution prior to infusion, catheter replacement, and in the most extreme cases transfer to hemodialysis2.
The ideal PD solution would have a neutral pH to prevent the occurrence of infusion pain and the need to manually inject sodium bicarbonate. In the present study, two new solutions with a physiological pH were evaluated. Both solutions have a pH 7.0 to 7.4, and in one solution the lactate buffer of conventional solutions has been replaced with a 38 mM bicarbonate buffer, and in the other solution a combined 25 mM bicarbonate/15 mM lactate buffer is utilized. The new solutions are presented in a two chambered bag configuration to separate calcium and magnesium from bicarbonate to avoid precipitation. The contents of the two chambers are mixed prior to administration. The aim of the present study was to assess the effectiveness of these two new solutions in alleviating inflow pain in patients who experienced such pain using conventional lactate containing PD solution using a randomized, double-blind, cross-over design.
METHODS
Participating centers
The study had a prospective, randomized, double-blind, cross-over design, where test and control solutions were studied during single dwells. A total of 18 patients were recruited from 8 participating centers (N is the number of patients recruited at each center): Stobhill Hospital, Glasgow, United Kingdom (N = 4); Manchester Royal Infirmary, United Kingdom (N = 5); Addenbrooke's Hospital, Cambridge, United Kingdom (N = 3); Royal Berkshire Hospital, Reading, United Kingdom (N = 1); Lasarettet, Hässleholm, Sweden (N = 2); Mälarsjukhuset, Eskilstuna, Sweden (N = 1); Länssjukhuset, Gävle, Sweden (N = 1); Universitetssjukhuset, Linköping, Sweden (N = 1). The study protocol was approved by the Ethics Committee of each participating center.
Study population
Patients were recruited to the study if they were at least 18 years old, gave fully informed consent to participate, and experienced repeated infusion pain that based on medical judgement was not related to the catheter or excessive intraperitoneal volume of dialysis fluid. Patients could not participate in the study if they had been treated for peritonitis within the previous 30 days.
Eighteen patients were recruited to the study. One patient withdrew due to transplantation during the study and was excluded from the statistical analysis. The mean age of patients at entry was 53.5 years, with a mean time on PD of 1.7 years and the mean length of time since first experiencing inflow pain was 16.4 months. Eleven of the 18 patients regularly used hypertonic glucose solution of at least 3.86%, and two patients commented that the infusion pain that they experienced was associated with the use of the 3.86% glucose solution (see below). There was a large variation within the eight participating centers in the frequency with which infusion pain was observed within each center's total CAPD population. The mean frequency across all centers was 1 in 25 (range 1 in 3 to 66) patients.
Half of the patients had experienced at least one episode of peritonitis in the previous six months. Two-thirds of the patients regularly injected alkalizing agents into their PD bags prior to infusion. The 12 patients who regularly injected alkalizing agents had a much higher incidence of peritonitis (1 episode/6 patient months), when compared with the 6 patients who did not regularly inject (1 episode/18 patient months). However, due to the small sample size this difference failed to reach statistical significance (P = 0.131, unpaired t-test).
Study procedures and test solutions
Patients attended the participating hospital dialysis clinic for all study exchanges. Within a one to three week period, patients were evaluated during two dialysis exchanges with each test solution in random order. Thus, all patients were to undergo six separate study dwells; patients could undertake a maximum of two test evaluations in one day, but it was required that these study exchanges were separated by a routine dwell (40 mM lactate solution) of at least four hours. All dwells were required to be of at least three hours and all evaluations were undertaken using a 3.86% glucose solution, as a previous study has suggested that infusion pain occurs more frequently with the use of hypertonic glucose solutions7. The composition of the three solutions is given in Table 1.
Due to the subjective nature of the pain response, a blinded procedure and a validated pain questionnaire were utilized. The identity of the solution under evaluation was unknown to the patient as well as the study nurse. The randomized schedule for solution administration for each patient was only known by an attending nurse who prepared and blinded the test bag prior to infusion (by the use of an opaque overpouch). No alkalizing agents (that is, manual injection of sodium bicarbonate) were permitted to be added prior to the infusion of any of the three solutions under investigation.
Pain assessment
Pain was assessed by two methods. A five point verbal scale (0 = no pain, 1 = mild pain, 2 = moderate pain, 3 = severe pain, 4 = very severe pain) was administered to assess pain intensity at the following time points: start of infusion, 1, 3, and five minutes after start of infusion, end of infusion, 10, 30 minutes and end of dwell, and start, five minutes and end of drain. Peak pain during the period of infusion, dwell and drain was defined as the most intense pain recorded with the five point verbal scale during each period.
In addition, the validated McGill Pain Questionnaire (MPQ)12,13 was completed by the patient after 40 minutes of the dwell. The MPQ was selected as the most suitable instrument for the assessment of inflow pain following a systematic search and review procedure, which identified 27 questionnaires that could potentially be suitable for use in the present study. The MPQ was selected because of its demonstrated validity, reliability and sensitivity, and the fact that it had previously successfully been used in a dialysis population14. The MPQ aims to capture not only the unidimensional aspects of pain related to the severity of the experience alone, as it now well established that pain is multidimensional and the severity or intensity of the sensory response is only one component. Thus, pain also has affective and evaluative dimensions which disrupt ongoing behavior and thought. Consequently, Melzack and Casey have postulated that there are three major dimensions of pain, namely the sensory-discriminative, motivational-affective, and cognitive-evaluative dimensions of pain15. All of these dimensions of pain are evaluated using the validated MPQ. Previous types of pain studies undertaken using the MPQ have shown that different types of pain (that is, headache, labor pain) have a unique profile in terms of their effects on the different MPQ dimensions and this has been used to discriminate between different types of pain15.
Safety was followed throughout the study by monitoring for adverse events.
Statistical analysis
Patients were included in the statistical analysis only if they completed the study (that is they received all of the six study exchanges). Including patients in the analysis who dropped out from the study could introduce a potential source of bias, as all patients would not receive the same number of study exchanges with each test solution. An initial analysis of variance (ANOVA) was undertaken with terms for patients, periods (the relative position of the six dwells), treatment and carryover16. If the carryover effect was significant at the P = 0.20 level, then an analysis was to be used to estimate and test the differences between the treatments. An analysis of variance (ANOVA) was used as it was considered the most appropriate methodology because it enabled estimates of treatment differences between the three test solutions to be made after allowing for all other factors in the design of the study. These other factors were: (1) differences between patients and participating centers; (2) differences between treatment periods (that is, the order of administration of the test solutions); and (3) any potential carryover effect by which the treatment given at one exchange might influence the results of subsequent evaluations. If no account were taken of these factors, for example, if a paired or unpaired t-test had been utilized, the results could be potentially misleading as the analysis would take no account of the factors above, which could potentially bias the results observed.
The actual study design was unbalanced due to one patient dropping out from the study and, therefore, the three solutions were not evaluated with equal frequency at each of the six single dwell positions. The results presented are the actual mean for each treatment adjusted after allowing for any possible differences as a result of the unbalanced administration. The adjusted mean was not significantly different from the actual mean for all variables assessed. Each patient was evaluated with each solution on two separate occasions and, therefore, the number of observations are exactly twice the number of patients analyzed. A P value less than 0.05 was considered statistically significant.
RESULTS
The pain severity time profile as assessed using the five-point verbal rating scale was typical of inflow pain Figure 1, with the most severe pain being observed during the first few minutes of infusion and then reducing in severity during the course of the dwell, and no pain being experienced during the later part of the dwell and drain.
Figure 1.
Time course of pain severity time profile in one patient during evaluation of the control lactate solution.
Full figure and legend (22K)The majority of patients experienced inflow pain with the control solution as assessed using the verbal rating scale Figure 2. The 17 evaluable patients were evaluated during two dialysis exchanges with each test solution in random order. All 17 evaluable patients met the entry criteria of experiencing repeated pain on infusion, but the pain response recorded with the control solution during the study showed a degree of variability in that no pain was recorded on 6 of 34 (18%) test evaluations upon infusion of the control lactate solution. Further, each patient was evaluated during two dialysis exchanges with the control solution in a random double-blind manner, and it is interesting to note that two patients experienced no pain during both exchanges, while a further two patients experienced no pain on only one of the exchanges with the control solution.
Figure 2.
Frequency of peak infusion pain severity using the verbal rating scale for the bicarbonate (
), bicarbonate/lactate (
) and lactate (
) solution. The 17 patients were evaluated with each solution on two separate occasions.
Both the bicarbonate (P = 0.001) and the bicarbonate/lactate (P = 0.001) solutions produced lower recorded peak pain during infusion compared to the lactate solution as assessed using the five-point verbal rating scale Figure 3.
Figure 3.
Peak pain verbal rating scale response (0 = no pain, 1 = mild pain, 2 = moderate pain, 3 = severe pain, 4 = very severe pain) for the lactate, bicarbonate and bicarbonate/lactate solution during (A) infusion, (B) dwell and (C) drain. Each condition represents the adjusted mean 
SEM for 34 separate evaluations of each solution. P values were calculated by ANOVA.
Pain of reduced intensity was also recorded in some patients during the dwell and drain periods Figure 3, despite the fact that all patients at entry met the inclusion criteria of experiencing repeated pain on infusion. Differences in the degree of pain reduction between the two bicarbonate containing solutions were also observed. The adjusted mean pain score for bicarbonate/lactate was lower than the corresponding score for the bicarbonate solution for peak pain during infusion, dwell and drain. This difference was statistically significant for peak pain during dwell (P = 0.048) and drain (P = 0.011).
Similar results for the effects of the study solutions on inflow pain were obtained with the MPQ Figure 4 and 5. Both the bicarbonate (P = 0.0002) and bicarbonate/lactate (P = 0.0001) solutions produced statistically significant lower scores for the total weighted pain rating index for the MPQ when compared to the control lactate solution. Furthermore, the total weighted pain rating index was lower for bicarbonate/lactate compared to the bicarbonate solution. Similar results were obtained for the individual MPQ dimensions (affective, evaluative, sensory, miscellaneous), the number of words chosen (which assesses pain severity) and the present pain intensity.
Figure 4.
Effects of the lactate, bicarbonate and the bicarbonate/lactate solutions on the (A) total weighted pain rating index, (B) number of words chosen and (C) present pain intensity of the McGill Pain Questionnaire. Each condition represents the adjusted mean SEM for 34 separate evaluations of each solution. P values were calculated by ANOVA.
Figure 5.
Effects of the lactate, bicarbonate and the bicarbonate/lactate solutions on individual dimensions of the McGill Pain Questionnaire: (A) affective dimension; (B) sensory dimension; (C) evaluative dimension; (D) miscellaneous dimension. Each condition represents the adjusted mean SEM for 34 separate evaluations of each solution. P values were calculated by ANOVA.
Following completion of each test evaluation, patients were asked to respond yes or no to the three questions shown in Table 2. Patient responses to all questions suggested that the intensity of the pain response was diminished with both the bicarbonate and bicarbonate/lactate solution compared to the control lactate solution. Furthermore, the patient responses to two of the three questions indicated that less severe pain was experienced with bicarbonate/lactate compared to the bicarbonate solution.
Two adverse events that were classified by the investigator as having a probable or possible relationship to the study solutions were recorded during the study. Both adverse events (general pruritus of mild severity and severe sweating) resulted from the administration of the control lactate solution.
DISCUSSION
Due to the subjective nature of the pain response, it was important in the present study to minimize any bias due to the patient or study nurse being aware of the identity of the solution being evaluated. This was overcome by the randomized, double-blind, cross-over design employed, and the fact that each patient was evaluated with each solution on two occasions. Furthermore, it is well established that infusion pain tends to diminish with time on PD3,6. The present study ensured that patients were studied in as short a time as reasonably possible and, by using a randomized design, no differences in effectiveness could be attributed to a reduction in pain intensity with time. Furthermore, a percentage of patients experienced no infusion pain when evaluated with the control lactate solution. Previous studies of new potential treatments for infusion pain have unfortunately not utilized such a vigorous design and, therefore, their conclusions should be interpreted with some caution8,10,11,17. Furthermore, it is recommended that all future studies assessing the effectiveness of new treatments for infusion pain utilize a similar vigorous design to that used in this study.
In the present study, both the bicarbonate and bicarbonate/lactate solution resulted in a significant reduction in infusion pain in patients who experienced such pain with conventional lactate buffered solution. These effects were demonstrated using both a pain severity rating index and the validated MPQ. There was a degree of variability in the pain response observed upon infusion of the control lactate solution, with the mean pain severity being intermediate between mild and moderate. This variability was observed between and within patients. The adjusted mean pain score for the total weighted pain rating index of the MPQ was similar to the recorded pain for the symptoms of cramps and bone pain previously studied in hemodialysis patients14. The pain severity experienced in the present study is likely to be an underestimate of the actual pain severity experienced in this patient population, as two of the patients were permitted for compassionate reasons (although a violation of the study protocol) to inject alkalizing agents to relieve intolerable pain following the infusion of the control lactate solution. It should be pointed out that due to the double-blind nature of the study both patients were unaware that the control solution was being evaluated. Pain, albeit of reduced severity, was also observed during the dwell and drain periods with the control solution, but only the bicarbonate/lactate solution produced a reduction in the pain intensity compared to both the control lactate and bicarbonate solution.
For all pain variables assessed with both the verbal rating scale and the MPQ, lower pain scores were seen with the bicarbonate/lactate compared to the bicarbonate solution, and for two of the variables assessed (peak pain during dwell and drain), this difference was statistically significant. This difference was supported by the patient responses to the additional questions asked at the end of each test evaluation. In addition, the bicarbonate solution appeared to result in more intense pain than the control lactate solution during the dwell and drain periods. The fact that bicarbonate/lactate would appear to be more effective than the bicarbonate solution in reducing inflow pain raises the possibility that such pain is not entirely due to pH as both solutions have identical pH. One possible explanation could be due to differences in the pCO2 of the two bicarbonate containing solutions. Indeed, pCO2 in the solutions was measured at 37°C by injecting a sample of freshly mixed solution into a blood gas analyzer, equipped with a pCO2 electrode and the pure bicarbonate solution had a significantly higher value than the combined solution Table 1. It is more difficult to explain why the bicarbonate solution would apparently result in more intense pain during dwell and drain than the conventional lactate solution. One explanation could be the high pCO2 level or the relatively high concentration (38 mM) of bicarbonate compared to a normal plasma level of approximately 25 mM18. Indeed, some aspects of cellular function are better preserved in vitro with bicarbonate/lactate than a pure bicarbonate solution19,20,2122.
In agreement with a previous report7, the present study demonstrated that peritonitis rates are higher in patients who manually inject alkalizing agents prior to infusion to alleviate pain. Although patients were only exposed to the novel solutions for a relatively short time interval, no episodes of peritonitis were observed following use of the bicarbonate solutions in the present study, and a previous two month evaluation in 47 patients has demonstrated that the bicarbonate containing solutions have similar peritonitis rates to conventional lactate containing solutions18.
In conclusion, both the bicarbonate and bicarbonate/lactate solution were effective in reducing the intensity of pain associated with the infusion of conventional lactate based PD solutions. The bicarbonate/lactate solution would appear to be more effective in alleviating inflow pain than the bicarbonate solution. This observed difference supports the hypothesis that the cause of inflow pain may be multifactorial and not just related to pH. Furthermore, the use of the bicarbonate/lactate solution for the treatment of inflow pain has the advantages over the most widespread current treatment, namely the manual injection of alkalizing agents, of not adding to the therapy burden of the patient and, most importantly, not increasing the risk of peritonitis.
References
| 1. | KHANNA R & OREOPOULOS DG. Complications of peritoneal dialysis other than peritonitis. Peritoneal Dialysis 1985; edited by NOLPH K Martinus Nijhoff, Boston pp 441−524. |
| 2. | TWARDOWSKI ZJ & KHANNA R. Peritoneal Dialysis Access and Exit Site Care. The Textbook of Peritoneal Dialysis 1994; edited by GOKAL R, NOLPH KD Berlin, Kluwer Academic Publishers pp 271−314. |
| 3. | PRAKASH J, TRIPATHI K & SINGH M. Complications of intermittent peritoneal dialysis. J Assoc Physicians India 1990; 38: 906−908. | PubMed | |
| 4. | CONN HO & FESSEL JM. Spontaneous bacterial peritonitis in cirrhosis: Variations on a theme. Medicine 1971; 50: 161−197. | PubMed | |
| 5. | PEDERSEN FB, RYTTOV N, DELEURAN P, DRAGSHOLT C & KILDEBERG P. Acetate versus lactate in peritoneal dialysis solutions. Nephron 1985; 39: 55−58. | PubMed | |
| 6. | VAAMONDE CA, MICHAEL VF, METZGER RA & CARROL KE. Complication of acute peritoneal dialysis. J Chronic Dis 1975; 28: 637−659 10.1016/0021-9681(75)90076-4. | PubMed | |
| 7. | HENDERSON IS, COUPER IA & LUMSDEN A. Potentially irritant glucose metabolites in unused CAPD fluid. inFrontiers in PD 1986; edited by MAHER JF, WINCHESTER JF New York, Field Rich and Associates pp 261−264. |
| 8. | YAMAMOTO T, SAKAKURA T, YAMAKAWA M, HORIUCHI N, HIRATA S, IRITANI Y, NISHITANI H, KIM S, KISHIMOTO T, CHIKU T, MATSUMOTO T & HORIO T. Clinical effects from long-term use of neutralised dialysate for continuous ambulatory peritoneal dialysis. Nephron 1992; 60: 324−329. | PubMed | |
| 9. | DAUGIRDAS JT, CHAUDHRY AW & ING IL. Alkalisation of peritoneal dialysis solutions with sodium hydroxide or sodium bicarbonate. Artif Organs 1988; 11: 203−205. |
| 10. | ROMAGNONI M, BECCARI M, FAIOLO S, GRANELLO E, SCALIA P & PALEARDI F. Abdominal pain with infusion of the peritoneal dialysis solutions relieved by alkalinization. Perit Dial Bull 1984; July-Sept: 188−189. |
| 11. | BUNCHMAN TE & BALLAL SH. Treatment of inflow pain by pH adjustment of dialysate in peritoneal dialysis. Perit Dial Int 1991; 11: 179−180. | PubMed | |
| 12. | MELZACK R. The McGill Pain Questionnaire: Major properties and scoring methods. Pain 1975; 1: 277−299 10.1016/0304-3959(75)90044-5. | Article | PubMed | ISI | ChemPort | |
| 13. | MELZACK R & KATZ J. The McGill Pain Questionnaire: Appraisal and current status. inHandbook of Pain Assessment 1992; edited by TURK DC, MELZACK R New York, Guilford Press pp 152−167. |
| 14. | BINIK YM, BAKER AG, KALOGEROPOULOS D, DEVINS GM, GUTTMANN RD, HOLLOMBY DJ, BARRE PE, HUTCHISON T, PRUD’HOMME M & MCMULLEN L. Pain, control over treatment, and compliance in dialysis and transplant patients. Kidney Int 1982; 21: 840−848. | PubMed | |
| 15. | MELZACK R & CASEY KL. Sensory, motivational and central control determinants of pain: A new conceptual model. inThe Skin Senses 1968; edited by KENSHALO D pp 423−439. |
| 16. | KIRKWOOD BR. Essentials of Medical Statistics 1988; London, Blackwell Science pp 47−56. |
| 17. | SIMONSEN O, WIESLANDER AP, LANDGREN C & RIPPE B. Less infusion pain and elevated levels of CA-125 by the use of a new and more bicompatible PD fluid. in KHANNA R, NOLPH KD, PROWANT BF, TWARDOWSKI ZJ, OREOPOULOS DG Advances in Peritoneal Dialysis, TorontoPerit Dial Bull 1996; 12: 156−160. |
| 18. | COLES GA, GOKAL R, OGG C, JANI F, O'DONOGHUE DT, CANCARINI GC, MAIORCA R, TRANAEUS A, FAICT D & DE VOS C. A randomized controlled trial of bicarbonate and a bicarbonate/lactate containing dialysis solution in CAPD. Perit Dial Int 1997; 17: 48−51. | PubMed | |
| 19. | SCHAMBYE HT, PEDERSEN FB & WANG P. Bicarbonate is not the ultimate answer to the biocompatibility problems of CAPD solutions: A cytotoxicity test of CAPD solutions and effluents. Adv Perit Dial 1992; 8: 42−46. | PubMed | |
| 20. | SCHAMBYE HT, PEDERSEN FB, CHRISTENSEN HK, BERTHELSEN H & WANG P. The cytotoxicity of continuous ambulatory peritoneal dialysis solutions with different bicarbonate/lactate ratios. Perit Dial Int 1993; 13 Suppl 2: S116−S118. | PubMed | |
| 21. | TOPLEY N, MACKENZIE R, WILLIAMS JD, MOSELY A, ARGYLE R, COLES GA, FAICT D, TRANÆUS A & HOLMES CJ. Acute in vivo exposure bicarbonate/lactate (TBL) and bicarbonate (TB) buffered peritoneal dialysis fluids (PDF) improves LPS driven peritoneal macrophage TNF  secretion. Perit Dial Int 1997; 17 Suppl 1: S39. |
| 22. | TOPLEY N, MACKENZIE R, WILLIAMS JD, COLES GA, FAICT D, TRANÆUS A, MOSELY A, ARGYLE R, JONES S & HOLMES CJ. Long-term in vivo exposure to bicarbonate-buffered PDF (TBL) improves ex vivo peritoneal macrophage function, comparison with bicarbonate (TB) or lactate buffered PDF (PD4). Perit Dial Int 1997; 17 Suppl 1: S39. |
Acknowledgments
The authors thank the study nurses for their assistance in conducting the study, and Ms. Noelle Alexandre and MetaData Solutions for database and statistical assistance.
