X. Appendices

Appendix A: Detailed Rationale for Guideline 1


When to Initiate Dialysis–Kt/Vurea Criterion (Opinion)

Unless certain conditions are met, patients should be advised to initiate some form of dialysis when the weekly renal Kt/Vurea (Krt/Vurea) falls below 2.0. The conditions that may indicate dialysis is not yet necessary even though the weekly Krt/Vurea is less than 2.0 are:

1. Stable or increased edema-free body weight. Supportive objective parameters for adequate nutrition include a lean body mass >63%, subjective global assessment score indicative of adequate nutrition (see Guideline 12: Nutritional Status Assessment, and Appendix B: Detailed Rationale for Guideline 2) and a serum albumin concentration in excess of the lower limit for the lab, and stable or rising; and

2. Nutritional indications for the initiation of renal replacement therapy are detailed in Guideline 27 of the NKF-K/DOQI Clinical Practice Guidelines on Nutrition, part of which are reproduced as Guideline 2 of the PD Adequacy Guidelines.

3. Complete absence of clinical signs or symptoms attributable to uremia.

A weekly Krt/Vurea of 2.0 approximates a renal urea clearance of 7 mL/min and a renal creatinine clearance that varies between 9 to 14 mL/min/1.73 m2. Urea clearance should be normalized to total body water (V) and creatinine clearance should be expressed per 1.73 m2 of body surface area. The GFR, which is estimated by the arithmetic mean of the urea and creatinine clearances, will be approximately 10.5 mL/min/1.73 m2 when the Krt/Vurea is about 2.0.

Rationale In patients with chronic kidney disease, progression of kidney failure should be monitored by following total weekly renal urea nitrogen clearance (Krturea) normalized to urea volume of distribution (V), ie, Krt/Vurea.1,2 This does not imply that a weekly collection of urine is necessary. A daily collection multiplied by seven yields a reasonable approximation of weekly clearance. The knowledge of Krt/Vurea is especially important when glomerular filtration rate (GFR) falls below 25 to 50 mL/min, at which time spontaneous decrease in dietary protein intake is commonly observed.3-6 The blood urea nitrogen (BUN) and serum creatinine values should not be used to monitor progression of renal failure, particularly in patients with diabetes.2 BUN may be low secondary to low protein intake and may not adequately reflect the degree of the renal functional impairment. Serum creatinine may be low due to decreased muscle mass as seen in some women, in the elderly, and in malnourished patients. Hence, serum creatinine concentration may not adequately reflect the degree of the renal functional impairment.

The estimation of V (total body water) by any formulae has not been validated in children with renal failure. Thus, the use of Krt/Vurea as an indication for the initiation of PD is recommended considering this caveat. Creatinine clearance as a means of assessing RRF for purposes of initiation of dialysis should be normalized to body surface area (BSA).

An increasing body of evidence1,7-10 suggests that Krt/Vurea is a reliable predictor of outcome in PD and that weekly values in the range of 2.0 provide adequate therapy (see Guideline 15: Weekly Dose of CAPD). Although the CANUSA data indicated a linear decrease in modeled mortality rate with increasing Kprt/V up to 2.3, there is some uncertainty about the significance of the high Kprt/V levels achieved in this study.11

It has always been a paradox that nephrologists have insisted on optimal therapy once patients are started on dialysis but have accepted much lower levels of renal function, defined as Krt/Vurea, during the predialysis phase of patient management. For example, while we recognize that a weekly Kprt/Vurea of 2.0 or higher is associated with improved outcome on PD, dialysis is usually not initiated until weekly Krt/Vurea is in the range of 0.71 to 1.3.2,12 It is possible that the consequences of delaying initiation of PD may be analogous to the experience of the National Cooperative Dialysis Study, wherein the mortality rate in the year after the study ended was more than twice as high in those randomized to the low dose dialysis protocols, even though they were returned to standard dialysis after completing 24 weeks in the high BUN (low clearance) arm of the study.13

The Work Group feels that the data of Ikizler et al,4 McCusker et al,12 and Pollock et al6 strongly demonstrate the linkage between decreasing kidney function and worsening nutritional status. In the CANUSA study, multiple estimates of nutritional status were associated with two estimates of RKF.12 Patients who started PD at lower levels of RKF had a worse nutritional status than those who started at a higher level of RRF. CANUSA further demonstrated an association between the relative risk of death and worse baseline serum albumin concentration, worse time-dependent SGA, and worse time-dependent percent lean body mass.1 While an association between risk of death and nPCR could not be demonstrated in the multivariate analysis, several univariate analyses did demonstrate an association of individual estimates of baseline nutritional status with survival.

These data are consistent with the observations of Bonomini et al14 who found that patients starting dialysis with a residual kidney creatinine clearance of <5 mL/min had a worse long-term outcome than patients starting incremental hemodialysis with a mean residual kidney creatinine clearance of 11 mL/min.

From these observational data, it seems reasonable to draw the following conclusions:

Once Krt/Vurea falls below 2.0 per week or creatinine clearance falls into the range of 9 to 14 mL/min/1.73 m2, initiation of dialysis or transplantation15,16 should be strongly advised. The patient should be considered to be at increasing risk with any further decreases in Krt/Vurea in the absence of renal replacement therapy intervention. If dialysis is not instituted when Krt/Vurea falls below 2.0, it is mandatory to document a negative inquiry for clinical signs or symptoms of uremia and that the following have not occurred: (a) more than a 6% involuntary reduction in edema-free usual body weight (%UBW) or to less than 90% of standard body weight (NHANES II) in less than 6 months; (b) a reduction in serum albumin by greater than or equal to 0.3 g/dL and to less than 4.0 g/dL (see Nutrition Guideline 3), in the absence of acute infection or inflammation, confirmed by repeat laboratory testing; or (c) a deterioration in SGA by one category (ie, normal, mild moderate, severe; see Nutrition Guideline 9 and Nutrition Appendix VI). When PD is initiated, the Kpt/Vurea could be increased incrementally17-21 such that the combined value of Krt/Vurea + Kpt/Vurea does not fall below the target level of 2.0 (see Fig II-1 and Guideline 15: Weekly Dose of PD). Alternatively, the initiation of "full dose" PD may be offered. Since residual kidney function (RKF) is such a crucial component of total solute removal utilizing the incremental initiation approach, more intense scrutiny of RKF is necessary. For initiation with full dose PD, less intense scrutiny of RKF is indicated. This is discussed in Guideline 3: Frequency of Delivered PD Dose and Total Solute Clearance Measurement Within Six Months of Initiation, and Guideline 5: Frequency of Measurement of Kt/Vurea, Total CCr, PNA, and Total Creatinine Appearance, which address frequency of measurements.

Fig II-1. Solute removal and initiation of PD. Example of tracking solute clearance measurements for a single patient over a 12-month period. The point at which incremental PD was initiated is indicated by the arrow.


In the CANUSA study,1 the weekly CCr equivalent to a Kprt/Vurea of 2.0 was 70 L/wk/1.73 m2. As will be clear later in this discussion, a CCr this high is indicative of residual kidney function, which was clearly present in the CANUSA patients at initiation of PD.

CAPD is the only continuous chronic renal replacement therapy with which to quantitatively compare continuous residual kidney solute clearance. The Work Group strongly supports the opinion that the outcome data for a weekly Kt/Vurea of 2.0 are so compelling that using the same figure for initiation of dialysis justifies the unknown but presumably small risks of performing peritoneal dialysis. Those risks include infections and the possibility that increasing the length of time on PD contributes to eventual patient "burn-out." If a patient is suspected to be at high risk for these complications, PD may not be the best choice for renal replacement therapy.

The Work Group recognizes that the patient will play a major role in accepting the initiation of dialysis based on a certain "laboratory value." It is the responsibility of the care providers to make clear to the patient the rationale for initiating dialysis when the above conditions become applicable. Reasons to justify a delay in initiating dialysis are listed above. These reasons should be documented, if present.

The Work Group also recognizes that for many clinicians, initiating dialysis based on Kt/Vurea is a new concept. Therefore, we have attempted to equate this to the traditional measure of urea clearance, creatinine clearance, and GFR (estimated by the arithmetic mean of urea and creatinine clearance).

What follows is an explicit quantitative approach to the concept of renal urea clearance (Kr urea, mL/min). We recommend that adequate PD be considered to require:

Kprt/V=2.0 per week (1)

where Kprt is total weekly peritoneal plus renal urea clearance. Guideline 15: Weekly Dose of CAPD, explains the rationale for recommending that Kprt/V be 2.0 per week.

For the purpose of this discussion, Kr urea is considered equivalent to Kp urea and, therefore,

Kprt/V=1.44*7*Kr urea/V, (2)

where 1.44 converts mL/min to L/day, when Krt/V is 2.0.

2.0=1.44*7*Kr urea/V and (3)

Kr urea=0.20*V (4)

Equation 4 shows that Kr urea must equal 0.2 times V when Krt/V = 2.0. It is of interest to compare this criterion to those developed independently22 for hemodialysis (HD). For twice weekly (biw) HD, a coefficient has been developed22 from urea kinetic modeling to convert Kr urea, mL/min, to L of equivalent urea clearance during each twice weekly dialysis. The twice weekly adequate level of Kt/Vurea23 in HD is 1.85 (double pool) and the coefficient to convert Kr urea to equivalent urea clearance during dialysis is 9.0 with units of L/treatment/mL Kr urea. Therefore,

1.85=9.0*Kr urea/V or (5)

Kr urea=0.20*V (6)

Equation 6 for biw HD is identical to Equation 4 for PD. For thrice weekly (tiw) HD the (double pool) coefficient previously developed22 is 5.0, therefore,

1.0=5.0*Kr urea/V or (7)

Kr urea=0.20*V (8)

Since Equations 4, 6, and 8 are identical, it is apparent that PD, biw HD, and tiw HD should all be started when Kr urea = 0.20*V. For an average patient with V = 35 L, this defines a level of Kr urea = 7.0 mL/min. There are constraints on the lower level of Kr urea for biw HD.22 As Fig II-1 suggests, treatment could be started incrementally once weekly Krt/Vurea falls below 2.0. Typically, this would involve a single overnight exchange, intended to restore Kt/Vurea to 2.0 per week. Ultrafiltration would not be needed since at this level of Krt/Vurea urine volumes are usually adequate.

Levels of Residual Renal Creatinine Clearance, Cr Cr, mL/min At Which Dialysis Should Be Initiated. There are no Kr Cr criteria for HD, and the criteria defining the contribution of residual kidney function (Kr Cr) to therapy are different from the criteria defining the contribution of peritoneal creatinine clearance (KpCr) to the dose of therapy. The problem arises from the argument that tubular secretion of creatinine should be subtracted from the total kidney creatinine clearance, and renal function with respect to creatinine clearance is best expressed as "GFR" as developed below. Therefore, the definitions used with respect to Kr Cr will each be considered separately and related to the level of Kt/Vurea for which Kr urea and Kp urea are considered simply additive.

In all instances, the total weekly CCr is normalized to 1.73 m2 of BSA. In order to compare this to the Kprt/Vurea, BSA must be normalized relative to V, eg, based on the Hume equations and taking the mean of the genders, 1.73 m2 = 35 L.24 It is also reasonable to extend this as a linear relationship over the domain of patient size although it should also be considered a gender-dependent relationship.

Consideration of Renal Contribution to Dose Expressed as Cr Cr. As noted above, in all cases the dose of total creatinine clearance is expressed as L/wk/1.73 m2 of BSA. The following will normalize total creatinine clearance to 1.73 m2 (nBSA) and Kt/Vurea to a standard V = 35 L corresponding to 1.73 m2 of BSA in order to develop constants relating the creatinine and urea-based dosage parameters. To the extent that BSA and V increase and decrease at the same ratio (reasonably valid), the constants developed are generalizable, and therefore, the relation between Kr Cr and nBSA can be expressed as follows:

[Kr Crt/nBSA]=1.44*7*Kr Cr=10.1•Kr Cr (9)

Equation 9 simply describes the total weekly liters of renal creatinine clearance as a function of Kr Cr and our normalized BSA of 1.73 m2.

For the Kt/Vurea normalized to V = 35, therefore,

Krt/nVurea=1.44*7*Kr urea/35=0.29*Kr urea (10)

and, therefore,

Kr urea=3.47*Krt/nVurea (11)


Kr Cr=2*Kr urea, (12)

we can substitute Equation 11 into Equation 12 to show

Kr Cr=2*3.47*Krt/nV=6.94*Krt/nVurea (13)

Substituting now Equation 13 into Equation 9 yields the following equation:

[Kr Crt/nBSA]=10.1[6.94(Krt/nV)]=70(Krt/nV) (14)

Equation 14 shows that if we define the renal contribution to PD therapy by Kr Cr, the level of weekly Kr Cr per 1.73 m2 must be 70 times the Krt/nVurea so at Krt/nVurea of 2.0, Kr Cr t/1.73 m2 is 140. For the average patient with V = 35 L and BSA = 1.73 m2, the required level of Kr Cr = 14 mL/min.

Consideration of Renal Contribution to Dose Expressed as "GFR." In this case the effective renal creatinine clearance, eKr Cr, is defined as:

eKr Cr=GFR=[Kr Cr+Kr urea]/2 (15)


[eKr Crt/nBSA]=1.44*7(Kr Cr+Kr urea)/2=5.0*Kr Cr+5.0*Kr urea (16)

Substituting from Equations 11 and 13, yields

[eKr Crt/nBSA]=5*6.94(Krt/V)+5*3.47(Krt/V)=52(Krt/V) (17)

Equation 17 shows that if the renal contribution to PD therapy is defined as GFR, which is equivalent to "effective" creatinine clearance or eKr Cr, the total weekly eKr Cr required relative to Krt/Vurea is 52 L/week/1.73 m2 per unit of Krt/Vurea. It can also be noted that in all instances these equations also relate to Kprt/Vurea and Kpt/Vurea since we have defined Kr = Kp.

Consideration of Peritoneal Creatinine Clearance to PD Dose. In this case, the dose must be expressed directly in terms of peritoneal creatinine clearance (KpCr) and by definition,

[KpCrt/nBSA]=1.44*7*KpCr=10.1 KpCr (18)

The average relationship between KpCr and Kp urea in CAPD is

KpCr=0.8*Kp urea (19)

Since Kp urea = Kr urea and Kp ureat/nV = Kr ureat/nV, Equations 18, 19, and 11 can be combined to derive

[KpCrt/nBSA]=28[Kp ureat/nV]=28[Kr ureat/nV] (20)

Equations 14, 17, and 20 show that the Creatinine Dose Equivalency with respect to the single urea Kprt /V criterion will vary widely depending on how RRF is defined. The relationships are plotted in Fig II-2: Dose of PD With Respect to Weekly Creatinine Clearance Relative to Weekly Kprt/V, where the weekly creatinine clearance required per 1.73 m2 corresponding to a weekly Kprt /Vurea of 2.0 ranges from 140 to 56 L, depending on the definitions used. There is no problem in either the case of pure residual kidney function with no PD therapy or the case of pure peritoneal dialysis with no residual kidney function present. The problem arises when one attempts to sum residual kidney creatinine clearance with peritoneal creatinine clearance. In this common circumstance, the dialysis dose relationship will be bounded by regression lines 2 and 3 in Fig II-2.

Fig II-2. Dose of PD with respect to weekly creatinine clearance relative to weekly Kprt/V. The dose of PD with respect to weekly creatinine clearance relative to weekly Kprt/V varies widely depending on the definition of renal creatinine clearance.


For line 1, which defines creatinine clearance as the uncorrected (for secretion) creatinine clearance, the creatinine clearance that is equivalent to a Kt/Vurea of 2.0 is 140 L/wk/1.73 m2. Line 2 defines creatinine clearance as the mean of urea and creatinine clearance, and the equivalence to a Kt/Vurea of 2.0 is 104 L/week/1.73 m2. Line 3 represents all clearance from PD (complete absence of residual kidney function). Under this condition a Kt/Vurea of 2.0 is equivalent to a creatinine clearance of 56 L/wk/1.73 m2.

This creates an irreconcilable ambiguity with respect to the creatinine and urea dosage criteria for defining optimal dialysis and the study of outcome as a function of dose. Because the practice of PD has used both CCr and Kt/V to quantify delivered dose and there is a large body of literature describing outcomes related to CCr, the Work Group recommends continuing to use both measures (see Guideline 4: Measures of PD Dose and Total Solute Clearance). However, in view of this ambiguity, the Work Group recommends that if only one measure is to be utilized, use Kprt/Vurea rather than Kr Crt (see Guideline 15: Weekly Dose of CAPD). Nonetheless, creatinine kinetics as discussed in Guidelines 4, 6, and 17 are useful for estimating edema-free, fat-free body mass, compliance with dialysis prescription, and some programs may prefer it for quantification of delivered dose of PD. Thus, total creatinine excretion is valuable.

Finally, it is worth emphasizing that the basic relationship between the level of residual Cr Cr to Kr urea for initiation of dialysis will be the same for all of these creatinine dosage criteria. They are related to Kprt/V since Kr Cr = 2 Kr urea and all of the expressions ultimately reduce to this relationship.

Another way to view the creatinine clearance at which to initiate dialysis is to extrapolate backward from the CAPD target of 60 L/week/1.73 m2 (see Guideline 15: Weekly Dose of PD). This approximates a purely filtered only creatinine clearance of 6 mL/min. Since at this level of residual renal function much of the creatinine appearing in the final urine is from tubular secretion, 60 L/wk/1.73 m2 approximates a total residual renal creatinine clearance of 9 to 14 mL/min.

Peritoneal Dialysis and Residual Kidney Function Equivalency. Quantitative replacement of renal urea clearance by peritoneal clearance is based on the assumption that the two clearance parameters confer equal clinical benefit with respect to control of uremic morbidity. Thus, we can write the relationship

Kprt/Vurea=Kpt/Vurea+Krt/Vurea, (21)

where Kpt/Vurea is total daily or weekly peritoneal urea clearance normalized to V; Krt/Vurea is total daily or weekly renal urea clearance normalized to V.

Solution of Equation 21 for Kpt/Vurea with the assumption that adequate weekly Kprt/Vurea is 2.0 results in

Kpt/Vurea=2.0- Krt/Vurea (22)

Equation 22 provides a quantitative guideline for replacing residual renal urea clearance by peritoneal clearance such that the sum of weekly Kpt/Vurea and Krt/Vurea remains 2.0.

The equivalence of peritoneal and residual renal clearance is controversial. Current data suggest inconsistent conclusions. There is a strong suggestion that protein metabolism in CAPD patients is similar to that in patients with chronic kidney failure.24 Peritoneal clearance has been shown to predict survival.25,26 However, a large retrospective analysis suggested that peritoneal clearance was not predictive of survival, while residual renal clearance was27 and indirect observations seem to corroborate that. Thus, the Work Group will continue to equate residual renal and peritoneal clearance. To that end, the preservation of residual renal clearance is paramount and strategies to achieve this have recently been described.28

Hemodialysis and Residual Kidney Function Equivalency. Compared to CAPD it is more complex to calculate incremental doses of HD such that the sum of intermittent dialyzer clearance (Kdt/Vurea) and continuous Krt/Vurea remain constant at a level equivalent to a weekly Krt/Vurea of 2.0. However, the dose and frequency of HD which provide therapy equivalent to continuous Kpt/Vurea can be calculated using the fundamental assumption underlying CAPD therapy: the level of continuous Kpt/Vurea required for treatment which is clinically equivalent to intermittent HD is that Kpt/Vurea which results in a steady state BUN equal to the average predialysis BUN with any specific intermittent HD treatment schedule at the same nPCR.29-32 From this basic assumption and the urea kinetic model19 the dose and frequency of HD required for incremental replacement of Krt/Vurea as it falls below 2.0 can be readily calculated, as depicted in Fig II-3.

The dose of intermittent HD is expressed in Fig II-3 as eKdt/Vurea, which is the equilibrated, delivered, and normalized hemodialysis dose (see the NKF-K/DOQI Clinical Practice Guidelines for Hemodialysis Adequacy for further discussion of eKdt/Vurea). The equilibrated measure is utilized here because in peritoneal dialysis, transcellular urea equilibration is achieved, and therefore, it makes conceptual sense to think in terms of equilibrated values. From preliminary data of the HEMO study, eKdt/V is approximately 0.21 lower than that computed from immediate post-HD BUN sampling, using single-pool, variable volume kinetic modeling.33 The dashed line depicts incremental increase in daily Kpt/Vurea as Krt/Vurea falls in accordance with Equation 22. Model solutions are shown for once, twice and thrice weekly hemodialysis (N = 1, 2, and 3, respectively). The model solutions are limited to eKdt/Vurea 2.0 since it is unrealistic to prescribe eKdt/Vurea >2.0. Such a dose would correspond to single pool Kdt/Vurea values of 2.8 and 2.3 with treatment times (t) of 2.0 and 4.0 hours, respectively. It can be seen that when N = 1, eKdt/Vurea = 2.0 when Krt/Vurea = 1.6. Thus, the option for once weekly hemodialysis is limited to a weekly Krt/Vurea 1.6. If Krt/Vurea = 0.5 and N =2, the eKdt/Vurea for each HD treatment must be 2.0 to achieve a weekly continuous Kt/Vurea equivalent to 2.0. Therefore, for a weekly Krt/Vurea <0.5, more than twice weekly HD will be necessary. Finally, in the case of N = 3, eKdt/Vurea increases linearly to 1.05 as Krt/Vurea falls to zero. The eKdt/Vurea of 1.05 corresponds to single pool Kdt/Vurea values of 1.46 and 1.20 at treatment times of 2.0 and 4.0 hours, respectively.

Fig II-3. Equivalent total dialysis doses for incremental replacement of Kprt/Vurea. Assumptions are made that the Kr, Kp, and Kd are clinically equivalent clearance items. Another assumption for this particular model is that an equal nPCR, the CAPD steady-state BUN equals the average prehemodialysis BUN. Thus, the intermittent hemodialysis dose can be related to the continuous dialysis (CAPD) by the various curves. N = (□) 1, ○ 2, or (■) 3 refers to once, twice, or thrice weekly hemodialysis treatments, respectively. The vertical axis is the equilibrated (double pool) delivered and normalized hemodialysis dose. Equilibrated Kt/V is about 0.21 lower than single pool and is necessary to use here because the CAPD steady-state is equilibrated.


There is emerging evidence that RRF is better preserved in patients undergoing HD with the use of more biocompatible membranes.34


1. Churchill DN, Taylor DW, Keshaviah PR: Adequacy of dialysis and nutrition in continuous peritoneal dialysis: Association with clinical outcomes. J Am Soc Nephrol 7:198-207, 1996

2. Tattersall JE: Is continuous ambulatory peritoneal dialysis an adequate long-term therapy for end-stage renal-disease? Semin Dial 8:72-76, 1995

3. Kopple JD, Chumlea WC, Gassman JJ, Hotlinger DL, Maroni BJ, Merrill D, Scherch LK, Schulman G, Zimmer GS: Relationship betweeen GFR and nutritional status–Results from the MDRD study. J Am Soc Nephrol 5:335A, 1994 (abstr)

4. Ikizler TA, Greene JH, Wingard RL, Parker RA, Hakim RM: Spontaneous dietary protein intake during progression of chronic renal failure. J Am Soc Nephrol 6:1386-1391, 1995

5. Hakim RM, Lazarus JM: Initiation of dialysis. J Am Soc Nephrol 6:1319-1328, 1995

6. Pollock CA: Protein intake in renal disease. J Am Soc Nephrol 8:777-783, 1997

7. Teehan BP, Schleifer CR, Brown JM, Sigler MH, Raimondo J: Urea kinetic analysis and clinical outcome on CAPD. A five year longitudinal study. Adv Perit Dial 6:181-185, 1990

8. Maiorca R, Cancarini G, Brunori G, Zubani R, Camerini C, Manili L, Campanini M, Mombelloni S: Which treatment for which patient in the future? Possible modifications in CAPD. Nephrol Dial Transplant 10:20-26, 1995

9. Tattersall JE, Doyle S, Greenwood RN, Farrington K: Kinetic modelling and underdialysis in CAPD patients. Nephrol Dial Transplant 8:535-538, 1993

10. Maiorca R, Brunori G, Zubani R, Cancarini GC, Manili L, Camerini C, Movilli E, Pola A, d’Avolio G, Gelatti U: Predictive value of dialysis adequacy and nutritional indices for mortality and morbidity in CAPD and HD patients. A longitudinal study. Nephrol Dial Transplant 10:2295-2305, 1995

11. Gotch FA, Gentile DE, Keen M: The incident of patient cohort study design with uncontrolled dose may result in a substantial overestimation of mortality (M) as a function of peritoneal dialysis (PD) dose. ASAIO Trans 42:102-102, 1996

12. McCusker FM, Teehan BP, Thorpe K, Keshaviah P, Churchill D: How much peritoneal dialysis is required for the maintenance of a good nutritional state? Kidney Int 50:S56-S61, 1996 (suppl 56)

13. Hakim R: Assessing the adequacy of dialysis. Kidney Int 37:822-832, 1990 - No Abstract Available

14. Bonomini V, Feletti C, Scolari MP, Stefoni S: Benefits of early initiation of dialysis. Kidney Int 28:S57-S59, 1985 (suppl 17) - No Abstract Available

15. Roake JA, Cahill AP, Gray CM, Gray DRW, Morris PJ: Preemptive cadaveric renal transplantation–Clinical outcome. Transplantation 62:1411-1416, 1996

16. Katz SM, Kerman RH, Golden DG, Grevel J, Camel S, Lewis RM, Van Buren CT, Kahan BD: Preemptive transplantation–An analysis of benefits and hazards in 85 cases. Transplantation 51:353-355, 1991

17. Keshaviah P, Emerson P, Nolph K: Timely initiation of dialysis: A urea kinetic approach. Am J Kidney Dis 33:344, 1999

18. Golper T: Incremental dialysis. J Am Soc Nephrol 9:S107, 1998

19. Burkhart JM, Jordan J: Initial clinical experience with timely/incremental dialysis. Perit Dial Inter 19:S51, 1999 (abstr, suppl 1)

20. Williams PF: Timely initiation of dialysis. Am J Kidney Dis 34:594, 1999 (letter)

21. Gotch FA, Keen M: Kinetic modeling in peritoneal dialysis, in Nissenson AR, Fine RN, Gentile DE (eds): Clinical Dialysis (ed 3). Norwalk, CT, Appleton and Lange, 1995, pp 343-375

22. Wish JB, Webb RL, Levin AS, Port FK, Held PJ: Medical appropriateness of initiating chronic dialysis in the U.S.: Results of the revised medical evidence report pilot study. J Am Soc Nephrol 5:345A, 1994 (abstr)

23. Tzamaloukas AH, Murata GH: Body surface area and anthropometric body water in patients on CPD. Perit Dial Int 15:284-285, 1995 (letter)

24. Mehrotra R, Saran R, Moore H, Prowant B, Khanna R, Twardowski Z, Nolph K: Toward targets for initiation of chronic dialysis. Perit Dialysis Int 17:497-508, 1997

25. Jager KI, Merkus M, Dekker F, Boeschoten E, Tijssen J, Stevens P, Bos WJ, Krediet R: Mortality and technique failure in patients starting chronic peritoneal dialysis: Results of the Netherlands Cooperative Study on the adequacy of dialysis. Kidney Int 55:1476-1485, 1999 - Full Text

26. Davies S, Phillips L, Russel G: Peritoneal solute transport predicts survival on CAPD independently of residual renal function. Nephrol Dial Transplant 13;962-968, 1998

27. Diaz-Buxo J, Lowrie E, Lew N, Zhang H, Zhu X, Lazarus JM: Associates of mortality among peritoneal dialysis patients with special reference to preitoneal transport rates and solute clearance. Am J Kidney Dis 33:523-534, 1999

28. Lamiere N, Van Biesen W: The impact of residual renal function on the adequacy of peritoneal dialysis. Perit Dialysis Int 17:S102-S110, 1997 (suppl 2)

29. Popovich RP, Moncrief JW: Kinetic modeling of peritoneal transport. Contrib Nephrol 17:59-72, 1979

30. Teehan BP, Schleifer CR, Sigler MH, Gilgore GS: A quantitative approach to the CAPD prescription. Perit Dial Bull 5:152-156, 1985

31. Keshaviah PR, Nolph KD, Prowant B, Moore H, Ponferrada L, Van Stone J, Twardowski ZJ, Khanna R: Defining adequacy of CAPD with urea kinetics. Adv Perit Dial 6:173-177, 1990

32. Gotch FA: Prescription criteria in peritoneal dialysis. Perit Dial Int 14:S83-S87, 1994

33. Eknoyan G, Levey AS, Beck GJ, Agodoa LY, Daugirdas JT, Kusek JW, Levin NW, Schulman G: The hemodialysis (HEMO) study: Rationale for selection of interventions. Semin Dial 9:24-33, 1996

34. McCarty JT, Jenson BM, Squillace BS, Williams AW: Improved preservation of residual renal function in chronic hemodialysis patients using polysulfone dialyzers. Am J Kidney Dis 29:576-583, 1997


Appendix B

The original PD Adequacy Guideline 2 and Appendix B have been replaced by Guideline 27 of the Nutrition Guidelines. Members of both the PD Adequacy and the Nutrition Work Groups developed the updated Guideline. The Guideline is reproduced as Guideline 2 of the updated PD Adequacy Guidelines without the reference citations which are given in the Nutrition guidelines. Therefore, the reader is encouraged to read the Nutrition Guidelines to obtain the references.

Appendix C: Detailed Rationale for Guideline 6


Assessing Residual Kidney Function (Evidence)

Residual kidney function (RKF), which can provide a significant component of total solute and water removal, should be assessed by measuring the renal component of Kt/Vurea (Krt/Vurea) and estimating the patient’s glomerular filtration rate (GFR) by calculating the mean of urea and creatinine clearance.

Rationale The contribution of RKF to total solute and water clearance is significant (30% to 50%), especially during the first few years of dialysis therapy. Assessment of RKF is important for several reasons. A substantial fraction (30%) of the total renal replacement therapy may be provided by RKF when a patient begins PD.1 After 2 years of PD, the RKF may still contribute about 15% of the total Kt/Vurea. Since the RRF contribution will be added to that of PD, it will be measured in the same units and for the same solutes.

Preservation of RKF may be of particular importance to the effectiveness of long term PD therapy. There is a progressive decline of RKF over time with both HD and PD. Several studies have compared the rate of decline of RKF with the two dialytic modalities2-8 and demonstrated that RKF is preserved better in patients undergoing PD therapies compared to HD. In a study of 25 CAPD patients and 25 HD patients, the rate of decline of creatinine clearance was significantly slower over the first 18 months of dialysis in the CAPD patients.2 The PD patients started dialysis with an uncorrected CCr of 4.4 mL/min, and after 18 months it was 4.0 mL/min. In the HD patients, CCr at initiation was 4.3 mL/min, and after 18 months it was 1.3 mL/min (P < 0.01 compared to PD).

Similar differences were observed in patients with diabetes.8 In another study comparing the urine output and CCr, the urine output dropped significantly in HD patients at the end of 1 year compared to 3 years in CAPD patients.3 The mean annual decline of CCr was identical in HD and CAPD for patients with primary glomerulopathy. However, in the groups with nephrosclerosis and tubulointerstitial nephritis, the rate of decline of CCr was significantly slower in CAPD compared to HD patients. In another retrospective study of 4 years duration which compared 55 CAPD patients to 57 HD patients, the rate of decline in the HD group was twice that of the CAPD group.4 This difference persisted after adjustment for age, gender, hypertensive status, and the use of ACE inhibitors. Children have a better preservation of urinary volume, but not GFR, in those receiving PD.7

Despite their limitations, these studies generally demonstrate a slower rate of decline of RKF in patients on PD compared to HD. They also demonstrate that the rate of the decline varies from patient to patient. The faster rate of decline of RKF in HD is speculated to be due to repetitive hypotensive episodes, possibly complement activation and cytokine release, and the possibility that the more efficient HD may remove GFR stimulatory factors.

Several methods for measuring the CCr component of RKF are available. These include the uncorrected CCr, a flat percentage of uncorrected CCr as an estimate of GFR, or the average of creatinine and urea clearance also as an estimate of GFR.

While each method has its particular merits, the Work Group recommends using the arithmetic mean of creatinine and urea clearances to determine the RKF component to CCr and as an estimate of GFR. Therefore, the CCr component of RKF will subsequently refer to residual renal CCr, corrected for secretion by taking the arithmetic mean of urea and creatinine clearances. This method was selected for several reasons. First, it was used in some of the major outcome studies used in establishing these guidelines (see Guideline 15: Weekly Dose of CAPD). Second, the corrected CCr correlates better with Kt/Vurea than the uncorrected CCr.9 Third, it makes conceptual sense because the peritoneal transport of creatinine is by diffusion and convection, not secretion. The correction process addresses this.

The MDRD study derived two equations which may approximate GFR,10 one utilizing demographic, serum, and urine variables:

GFR in mL/min per 1.73 m2= 198

× (serum creatinine concentration in mg/dL)- 0.858

× (age)- 0.167_(0.822 if patient is female)

× (1.178 if patient is black)

× (serum urea nitrogen in mg/dL)- 0.293

× (urine urea nitrogen in g/day)+0.249

and the other equation using demographic and serum variables only:

GFR in mL/min per 1.73 m2=170

× (serum creatinine concentration in mg/dL)- 0.999

× (age)- 0.176_(0.762 if patient is female)

× (1.180 if patient is black)

× (serum urea nitrogen in mg/dL)- 0.170

× (serum albumin concentration in g/day)+0.318

However, the Work Group recommends using urea clearance, normalized to total body water, ie, Krt/Vurea, as the key measure to follow serially to determine whether urine collections need to continue (see Guideline 11: Dialysate and Urine Collections). This is termed the renal Kt/Vurea or Krt/Vurea. This recommendation was made to simplify the concept of a residual kidney component to the total renal replacement dose. Kt/Vurea is believed to be the more valuable measure of renal replacement therapy and the Work Group carried this thinking through to using Krt/Vurea in both initiation of dialysis and for following RKF changes over time.

Krt/Vurea as a measure of RKF is recommended because total Kt/Vurea is associated in a clinically important and statistically significant way with patient survival1,11,12 (see Guideline 15: Weekly Dose of CAPD). The peritoneal clearance of creatinine is about 80% of the urea clearance, while at end-stage the kidney clearance of creatinine is about 1.5 to 2 times that of urea. Perhaps as a consequence of this physiological phenomenon or for other reasons, there is a discrepancy between total Kt/Vurea and total CCr normalized to 1.73 m2 BSA (see paragraphs below). In the case of discrepancy, Kt/V urea is preferentially recommended to determine PD adequacy, because it is more predictable and reproducible and is independent of the confounding effects of renal secretion of creatinine. A retrospective study of PD adequacy demonstrated an association between Kt/Vurea and outcomes.13

This emphasis on using Krt/Vurea is not intended to detract from the utility of CCr. In terms of validity, total CCr normalized to 1.73 m2 BSA is predictive of patient survival, technique survival, and hospitalization.1 The creatinine generation rate is useful for assessment of nutritional status, in particular, in measuring fat-free, edema-free body mass. Total CCr may also be useful for assessment of compliance.

CCr as an index of PD adequacy is associated statistically with both morbidity and mortality1 and correlates with urea clearance.14 Discrepancies between the two clearances may be found in approximately 20% of PD subjects.15,16 The main reasons for the discrepancies are the presence of substantial RKF, which tends to cause disproportionately high CCr values, and low peritoneal solute transport type, which tends to cause disproportionately low CCr values.15,16 CCr values corresponding to a weekly Kt/Vurea of 2.0 differ between CAPD subjects with and without RKF (see Fig II-2, referenced previously in Appendix A: Detailed Rationale for Guideline 1). In patients with RKF the mean CCr corresponding to a Kt/Vurea of 2.0 weekly is between 60.514 and 67.615 L/wk/1.73 m2. In anuric CAPD subjects, the mean CCr corresponding to a Kt/Vurea of 2.0 weekly is 52.1 L/wk/1.73 m2.16

In the case of a discrepancy, the Work Group recommends the use of Kt/Vurea as an immediate guide of dialysis adequacy because it directly relates to protein metabolism. However, if there is a discrepancy between CCr and Kt/Vurea, the patient must be observed closely because initially it may not be clear why the discrepancy exists and the reason may be important. This is discussed in Guidelines 1, 7, and 15.


1. Churchill DN, Taylor DW, Keshaviah PR: Adequacy of dialysis and nutrition in continuous peritoneal dialysis: Association with clinical outcomes. J Am Soc Nephrol 7:198-207, 1996

2. Rottembourg J, Issad B, Gallego JL, Degoulet P, Aime F, Gueffaf B, Legrain M: Evolution of residual renal function in patients undergoing maintenance haemodialysis or continuous ambulatory peritoneal dialysis. Proc Eur Dial Transplant Assoc 19:397-403, 1982

3. Cancarini GC, Brunori G, Camerini C, Brass S, Manili L, Maiorca R: Renal function recovery and maintenance of residual diuresis in CAPD and hemodialysis. Perit Dial Bull 6:77-79, 1986

4. Lysaght MJ, Vonesh EF, Gotch F, Ibels L, Keen M, Lindholm B, Nolph KD, Pollock CA, Prowant B, Farrell PC: The influence of dialysis treatment modality on the decline of remaining renal function. ASAIO Trans 37:598-604, 1996

5. Hallett M, Owen J, Becker G, Stewart J, Farrell PC: Maintenance of residual renal function: CAPD versus HD. Perit Dial Int 12:S46A, 1992 (abstr)

6. Lutes R, Perlmutter J, Holley JL, Bernardini J, Piraino B: Loss of residual renal function in patients on peritoneal dialysis. Adv Perit Dial 9:165-168, 1993

7. Feber J, Scharer K, Schaefer F, Mikova M, Janda J: Residual renal function in children on haemodialysis and peritoneal dialysis therapy. Pediatr Nephrol 8:579-583, 1994

8. Rottembourg J: Residual renal function and recovery of renal function in patients treated by CAPD. Kidney Int 43:S106-S110, 1993 (suppl 40)

9. Bhatla B, Moore HL, Nolph KD: Modification of creatinine clearance by estimation of residual creatinine and urea clearance in CAPD patients. Adv Perit Dial 11:101-105, 1995

10. Levey AS, Bosch J, Breyer LJ, Greene T, Rogers N, Roth D: A more accurate method to estimate glomerular filtration rate from serum creatinine: A new prediction equation. Ann Intern Med 130:461, 1999

11. Maiorca R, Brunori G, Zubani R, Cancarini GC, Manili L, Camerini C, Movilli E, Pola A, d’Avolio G, Gelatti U: Predictive value of dialysis adequacy and nutritional indices for mortality and morbidity in CAPD and HD patients. A longitudinal study. Nephrol Dial Transplant 10:2295-2305, 1995

12. Teehan BP, Schleifer CR, Brown JM, Sigler MH, Raimondo J: Urea kinetic analysis and clinical outcome on CAPD. A five year longitudinal study. Adv Perit Dial 6:181-185, 1990

13. Selgas R, Bajo MA, Fernandez-Reyes MJ, Bosque E, Lopez-Revuelta K, Jimenez C, Borrego F, De Alvaro F: An analysis of adequacy of dialysis in a selected population on CAPD for over 3 years: The influence of urea and creatinine kinetics. Nephrol Dial Transplant 8:1244-1253, 1993

14. Nolph KD, Twardowski ZJ, Keshaviah PR: Weekly clearances of urea and creatinine on CAPD and NIPD. Perit Dial Int 12:298-303, 1992

15. Chen HH, Shetty A, Afthentopoulos IE, Oreopoulos DG: Discrepancy between weekly KT/V and weekly creatinine clearance in patients on CAPD. Adv Perit Dial 11:83-87, 1995

16. Tzamaloukas AH, Murata GH, Malhotra D, Fox L, Goldman RS, Avasthi PS: Creatinine clearance in continuous peritoneal dialysis: Dialysis dose required for a minimal acceptable level. Perit Dial Int 16:41-47, 1996

Appendix D: Detailed Rationale for Guideline 8


Reproducibility of Measurement (Opinion)

Accurate measurement of total Kt/Vurea and total creatinine clearance (Ccr) requires collection and analysis of urine, dialysate, and serum in a way that yields reproducible and valid results. Dialysate creatinine concentration must be corrected for the presence of glucose in some assays. Peritonitis precludes reliable measurement of delivered PD dose for up to 1 month. Compliance with complete collections is mandatory. For patients who void 3 times per day, a 24-hour urine collection is sufficient. For patients who void less frequently, a 48-hour collection is recommended. For CAPD patients, the serum sample can be obtained at any convenient time. For NIPD patients, the serum sample should be obtained at the midpoint of the daytime empty period. For CCPD patients, the serum sample should be obtained at the midpoint of the daytime dwell(s).

Rationale To be clinically useful, measurement of PD dose must be performed in a valid and reproducible fashion. The following factors influence the validity and reliability of Kt/Vurea and total CCr as measures of PD dose.

Dialysate glucose. Dialysate creatinine concentration should be corrected for the presence of glucose, which interferes with some creatinine measurement methodologies.1 Each facility must determine this by specifically inquiring of its laboratory whether the creatinine assay used by that lab is altered by high glucose concentrations. Each laboratory should establish its own correction factor and should reestablish the correction factor if the laboratory’s methodology changes. The Work Group does not recommend using correction factors from the literature.

Peritonitis. Peritoneal solute transport increases during peritonitis and usually recovers some time after resolution of peritonitis, with a reported recovery time between 3 days2 and 1 month.3

Patient compliance with the dialysis prescription. Following creatinine appearance in dialysate and urine longitudinally is an objective method to evaluate the degree of compliance (see Guideline 7: PD Dose Troubleshooting). Clinical tools for evaluating compliance are in the process of development.4

Variability of residual renal function (RRF). Day-to-day total clearances can vary greatly in PD. The major portion of this variance is caused by changes in measured RKF,5 although creatinine generation may vary in apparently stable PD patients.6

Completeness of urine collection. To avoid sampling errors, urine should be collected over 48 hours in patients who void infrequently (<3 times in 24 hours). A 24-hour urine sample can be used for all other patients. The urine collection should be performed on the same day as the dialysate collection. In children, the urine collection period may be reduced to a minimum of 12 hours.

Dialysate and urine collection for PD adequacy studies. Two methods of dialysate sampling predominate. In the first method, for CAPD, all effluent bags in a 24-hour period are brought to the center. While this "batch" method is simple in concept, it is difficult to carry out because it means transporting all the dialysate bags, which are heavy and bulky.

The second method is referred to as the "aliquot" method. In this approach, each bag of effluent dialysate is shaken vigorously for a few seconds, then is emptied into a measuring container accurate to an error of <50 mL per 2,000 mL. The volume for that bag is recorded in mL and the decimal point is moved three places to the left. The resulting figure is the number of mL which must be drawn from the dialysate effluent in the measuring container and placed in the laboratory red top test tube, provided by the dialysis center. For example, if the effluent volume for the CAPD bag is 2,450 mL, moving the decimal point three places to the left means that 2.45 mL of this fluid is put in the test tube (or other small collection container). Each bag for the 24-hour interval is handled this way. The aliquots are measured by syringes; usually a 5-mL syringe is accurate enough. The aliquots can be mixed in the same container, because the sampling proportion from each original bag is constant at 1/1,000. The total effluent is recorded, and that figure plus the small container with all the collected aliquots are brought to the dialysis center. Some dialysate manufacturers have developed special aliquoting exchange bags that separate an aliquot as part of the exchange.7

The collection of effluent dialysate from automated PD is conceptually similar to that described above. The effluent drained via a cycler is quantified automatically by the cycler and generally pools in one collection container or, if in several containers, free mixing is possible. Since the effluent volume is known and the containers are allowed to mix freely, a sample of any reasonable volume (eg, 10 mL) can be brought to the dialysis unit. If several containers are used with equal filling, an equal volume aliquot from each container can be pooled. The total effluent volume must be known and recorded. If the effluent bags are not freely mixing, then a sample from each bag is required, as well as the exact volume of the container from which the sample was drawn. One cannot extrapolate from one container (bag) to the next.

No matter what method is used for dialysate, a complete and accurately timed urine collection is necessary. The urine volume is more easily managed since it is much smaller than the dialysate volume. The longer the collection interval, the more reliable are the collections, assuming patient compliance. A timed collection of 12 to 48 hours is recommended, depending on how frequently the patient voids. Polyuric patients, particularly children, or patients with a short attention span, may void frequently enough that a supervised 12-hour collection is accurate. As for any urine collection, the bladder should be emptied and the urine discarded moments before the start of the collection period. Then, moments before the end of the collection period, the patient empties the bladder, and this urine, plus all that has been collected in the interval, completes the collection. The patient should try to delay the final voiding until just before the interval ends. Three or more bladder voidings generally are necessary for urine collections. For patients who make little urine and hence void infrequently, 48-hour collections may be more informative.

Serum samples. In CAPD, serum concentrations of urea and creatinine are relatively constant, and thus blood samples can be drawn at any convenient time for clearance determinations. For the asymmetric therapies (NIPD and CCPD), blood concentrations are lowest at the end of the cycling period and highest prior to the next cycling period. Theoretically, the best time to draw these blood samples is half way between the lowest and highest times. For NIPD patients, the serum sample should be obtained at the midpoint of the daytime empty period. For CCPD patients, the serum sample should be obtained at the midpoint of the daytime dwell. For most NIPD and CCPD patients, this time point conveniently occurs in the early afternoon.


1. Farrell SC, Bailey MP: Measurement of creatinine in peritoneal dialysis fluid. Ann Clin Biochem 28:624-625, 1991 - No Abstract Available

2. Raja RM, Kramer MS, Rosenbaum JL, Bolisay C, Krug M: Contrasting changes in solute transport and ultrafiltration with peritonitis in CAPD patients. Trans Am Soc Artif Intern Organs 27:68-70, 1981 - No Abstract Available

3. Krediet RT, Zuyderhoudt FMJ, Boeschoten EW, Arisz L: Alterations in the peritoneal transport of water and solutes during peritonitis in continous ambulatory peritoneal dialysis patients. Eur J Clin Invest 17:43-52, 1987

4. Bernardini J: Predicting compliance in home peritoneal dialysis (PD) patients. ASAIO J 42:102A, 1996 (abstr)

5. Rodby RA, Firanek CA, Cheng YG, Korbet SM: Reproducibility of studies of peritoneal dialysis adequacy. Kidney Int 50:267-271, 1996

6. Johansson A-C, Attman P, Haraldsson B: Creatinine generation rate and lean body mass: A critical analysis in peritoneal dialysis patients. Kidney Int 51:855-859, 1997

7. Smith L, Folden T, Youngblood B, Panlilio F, Gotch F: Clinical evaluation of a peritoneal dialysis kinetic modeling set. Adv Perit Dial 12:46-48, 1996

Appendix E: Detailed Rationale for Guideline 9


Estimating Total Body Water and Body Surface Area (Opinion)

V (total body water) should be estimated by either the Watson1 or Hume2 method in adults using actual body weight, and by the Mellits-Cheek method3 in children using actual body weight.

Watson method1:

For Men: V(liters) = 2.447 + 0.3362*Wt(kg) + 0.1074*Ht(cm) - 0.09516•Age (years)

For Women: V = - 2.097 + 0.2466*Wt + 0.1069•Ht

Hume method2:

For Men: V = - 14.012934 + 0.296785*Wt + 0.194786*Ht

For Women: V= - 35.270121 + 0.183809*Wt + 0.344547*Ht

Mellits-Cheek method for children3:

For Boys: V (liters)= - 1.927 + 0.465*Wt(kg) + 0.045*Ht(cm), when Ht < 132.7 cm

V= - 21.993 + 0.406*Wt + 0.209*Ht, when height is > 132.7 cm

For Girls: V=0.076 + 0.507*Wt + 0.013*Ht, when height is < 110.8 cm

V=- 10.313 + 0.252*Wt + 0.154*Ht, when height is > 110.8 cm

Body surface area (BSA) should be estimated by either the DuBois and DuBois method,4 the Gehan and George method,5 or the Haycock method6 using actual body weight.

For all formulae, Wt is in kg and Ht is in cm:

DuBois and DuBois method: BSA(m2)=0.007184*Wt0.425*Ht0.725

Gehan and George method: BSA(m2)=0.0235*Wt0.51456*Ht0.42246

Haycock method: BSA(m2)=0.024265*Wt0.5378*Ht0.3964

Rationale The practical methods described in the literature to estimate V include a fixed fraction of body weight (0.60 in males and 0.55 in females, or 0.58 in all subjects) and anthropometric formulae based on sex, age, height, and weight.1-3 The fixed fraction method is inaccurate, as it overestimates total body water even in overhydrated PD subjects.7 Therefore, the Work Group recommends that this method not be used.

Both the Watson1 and Hume2 formulae were derived by comparing anthropometric measurements to measurements of body water by indicator dilution techniques. An advantage of these formulae is that they were derived in populations which included obese subjects and, therefore, can account for obesity. Estimates from the Watson and Hume formulae are, in general, close to isotopic body water measurements in PD patients.7 The error of the estimates based on these formulae can increase in subjects with abnormalities in body water (hydration), because such subjects were systematically excluded from the studies used to derive the formulae.8 Indeed, both the Watson and Hume formulae tend to consistently underestimate isotopic body water in both lean and obese PD patients with overhydration.7 The formulae are recommended as a reasonable approximation with systematic error, but acceptable based on ease of determination. A proposed correction of the formulae for edema requires careful assessment of the dry weight. The correction considers actual weight at the edematous state as the sum of two weights: the dry weight plus the weight gain secondary to edema. V is the body water at dry weight obtained from the Watson or Hume formulae plus, in its totality, the weight gain secondary to edema.8 The Watson and Hume formulae provide similar estimates of V.9 Both formulae provide unrealistic estimates of V in subjects whose height and/or weight differ greatly from the ordinary.9 The Mellits-Cheek formulae were derived from subjects aged 1 month to 34 years for males and 1 month to 31 years for females. In each case, the measurement of total body water was performed in normal subjects by the use of deuterium oxide distribution with simultaneous measurement of weight and height.3

Body surface area is derived from anthropometric variables.4-6 While the formulae were derived in normal subjects, the influence of clinical conditions on the variability of the calculations are much less than that noted for total body water calculations.

Unfortunately, the relationship between the calculations for V and BSA is not linear.10 When BSA increases linearly as obesity develops, V increases exponentially. There are gender differences in these relationships as well. For the same height and BSA, males have a larger V than females. Future investigations should apply the same size indicator normalization factor to both solutes.

The Work Group considered the special case of the malnourished, underweight patient. Severe malnutrition is associated with low levels of RKF in patients who did not increase the dose of PD to compensate for the loss in renal function.11 This evidence suggests that the loss in kidney function may have caused underdialysis as Kprt/Vurea decreased. Underdialysis then caused uremia with anorexia and weight loss, which, in turn, resulted in lower V and higher Kprt/Vurea. Thus, while Kprt/Vurea may be in the "acceptable" range in underweight, malnourished individuals, improved nutrition and weight gain is of paramount importance in these individuals. If this aim is fulfilled, V will increase and Kprt/Vurea will decrease to the previous levels, which were inadequate. Therefore, it is recommended in such individuals to provide a dose of PD which will result in adequate Kprt/Vurea when they reach their desired weight without changing the dialysis prescription. For malnourished patients defined by SGA or Table II-3 below, provide a PD dose to achieve a weekly Kt/Vurea of 2.0 for the volume of the patient at desired weight. The calculation of the target Kprt/Vurea in malnourished CAPD subjects equivalent to a weekly Kprt/Vurea of 2.0 at desired weight is as follows: If Vactual is body water obtained from the Watson or Hume formulae using the actual weight and Vdesired is body water obtained by the same formulae using the desired weight, then for malnourished subjects Vactual < Vdesired. The target CAPD weekly Kprt/Vdesired is 2.0 (for CCPD 2.1 and NIPD 2.2). If Kprt/Vactual = x, then for the CAPD patient:

(Kprt/Vdesired)/(Kprt/Vactual)=2.0/x or (1)

2.0*Vdesired/Vactual=x, and x is the new target Kt/Vurea (2)

Equation 2 can be used to calculate the target Kprt/Vurea in malnourished CAPD subjects. For example, if Vactual is 35 L and Vdesired is 40 L, then the weekly target Kprt/Vactual is 2.0 times 40/35 or 2.29. In essence, the target of 2.0 is modified upward by a factor of Vdesired/Vactual. The recommendation to increase the target Kprt/Vurea in malnourished PD subjects is based on indirect evidence.

The above water volume estimations are used in the Kt/Vurea measure. Target Kt/Vurea is modified (Equation 2). The same principle applies if CCr is used. The normalization of CCr by BSA should correct weightactual for weightdesired in the formula used to determine BSA. This is further discussed in Guideline 15: Weekly Dose of CAPD.

The concept above is intended to deliver a dose of PD considered adequate for the patient at a "desired" weight. Defining a "desired" weight can be subjective, but objective definitions are available. One preferable method is that proposed by a broad collaborative "glossary" group where the "desired" weight described in this rationale is the glossary group’s "normal weight," defined as the median body weight of normal Americans with the same age, height, sex, and skeletal frame as the patient in question.12 Table II-3 from the glossary details these weight ranges based on the input parameters of the patient in question.

Alterations Caused by Amputation. The anthropometric formulae for total body water calculate that as obesity develops and body weight increases, V also increases, but body water content (the ratio V/weight) decreases. This is consistent with the known fact that water content of fat tissue is low. Calculations of V in amputees by uncorrected anthropometric formulae (using the actual postamputation weight and height in the calculations) distorts the relationship between V and weight. In this case, body water content is not consistent with the degree of obesity.13,14 The anthropometric formulae can be corrected in a way that restores the relationship between body water content and degree of obesity in three steps:

Step A: The fraction of body weight lost to amputation (fw) is obtained from a nomogram15 (see Table II-4). The fraction of weight loss (fw) is the percent loss in weight from Table II-4 divided by 100. The hypothetical nonamputated weight at the same body composition would be equal to actual weight/(1 - fw).

Step B: V at the hypothetical nonamputated weight (Vnon-amputated) is calculated from the Watson formula. Body water content is Vnon-amputated/Weightnon-amputated.

Step C: Actual V is calculated by multiplying the actual postamputation weight by Vnon-amputated/Weightnon-amputated. This correction makes the assumption that amputation per se will not change body water content.14

The calculations for body surface area (BSA) in amputees should be also corrected, because of inconsistent results obtained with the uncorrected calculation of BSA.16 The correction requires also three steps:

Step A: Same as step A in the correction of V in amputees.

Step B: BSA at the hypothetical nonamputated weight is calculated by one of the three BSA formulae above.

Step C: The fraction of BSA corresponding to the amputated limb(s) (fBSA) is obtained from Table II-4 derived from Herndon. The fraction (fBSA) lost is the percent loss from Table II-4 divided by 100. Corrected BSA is BSAnon-amputated times (1 - fBSA).17


1. Watson PE, Watson ID, Batt RD: Total body water volumes for adult males and females estimated from simple anthropometric measurements. Am J Clin Nutr 33:27-39, 1980

2. Hume R, Weyers E: Relationship between total body water and surface area in normal and obese subjects. J Clin Pathol 24:234-238, 1971 - No Abstract Available

3. Mellits ED, Cheek DB: The assessment of body water and fatness from infancy to adulthood. Monographs Soc Res Child Dev, Serial 140 35:12-26, 1970

4. Du Bois D, Du Bois EF: A formula to estimate the approximate surface area if height and weight be known. Nutrition 5:303-311, 1989 - No Abstract Available

5. Gehan E, George SL: Estimation of human body surface area from height and weight. Cancer Chemother Rep Part I 54:225-235, 1970

6. Haycock GB, Chir B, Schwartz GJ, Wisotsky DH: Geometric method for measuring body surface area: A height-weight formula validated in infants, children, and adults. J Pediatr 93:62-66, 1978

7. Wong KC, Xiong DW, Kerr PG, Borovnicar DJ, Stroud DB, Atkins RC, Strauss BJG: Kt/V in CAPD by different estimations of V. Kidney Int 48:563-569, 1995

8. Tzamaloukas AH: Effect of edema on urea kinetic studies in peritoneal dialysis patients. Perit Dial Int 14:398-401, 1994 - No Abstract Available

9. Tzamaloukas AH, Dombros NV, Murata GH, Nicolopoulou N, Dimitriadis A, Kakavas J, Malhotra D, Antoniou S, Balaskas EV, Voudiklari S: Fractional urea clearance estimates using two anthropometric formulas in continuous peritoneal dialysis: Gender, height and body composition differences. Perit Dial Int 16:135-141, 1996

10. Tzamaloukas A, Malhotra D, Murata G: Gender degree of obesity and discrepancy between urea and creatinine clearance in peritoneal dialysis. J Am Soc Nephrol 9:497, 1998

11. Jones MR: Etiology of severe malnutrition: Results of an international cross-sectional study in continuous ambulatory peritoneal dialysis patients. Am J Kidney Dis 23:412-420, 1994

12. Kopple JD, Jones MR, Keshaviah PR, Bergstrom J, Lindsay RM, Moran J, Nolph KD, Teehan BP: A proposed glossary for dialysis kinetics. Am J Kidney Dis 26:963-981, 1995

13. Tzamaloukas AH, Saddler MS, Murphy G, Morgan K, Goldman RS, Murata GH, Malhotra D: Volume of distribution and fractional clearance of urea in amputees on continuous ambulatory peritoneal dialysis. Perit Dial Int 14:356-361, 1994

14. Tzamaloukas AH, Murata GH: Estimating urea volume in amputees on peritoneal dialysis by modified anthropometric formulas. Adv Perit Dial 12:143-146, 1996

15. Hopkins B: Assessment of nutritional status. Nutritional Support Dietetics Corte Curriculum. (ed 2). Silver Spring, MD, American Society for Parenteral Nutrition, 1993, pp 15-70

16. Tzamaloukas AH, Malhotra D: Creatinine clearance in amputees on CPD. Perit Dial Int 16:426 1996

17. Anonymous: Pre-hospital management, transportation and emergency care, in Herndon DN (ed): Total Burn Care. Philadelphia, PA, Saunders, 1996, pp 36-38

18. US Renal Data Systems: Tthe USRDS Dialysis Morbidity and Mortality Study (Wave 1), in National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases (ed): U.S. Renal Data Systems 1997 Annual Data Report. Chapter 4. Bethesda, MD, 1997, pp 49-67

19. Frishanco AR: Anthropometric Standards for the Assessment of Growth and Nutritional Status. Ann Arbor, MI, University of Michigan Press, 1990

Table II-3.

Median (50th Percentile) Body Weight (kg)

Age Range Males Females
18.0-24.9 Frame index <38.4 38.4-41.6 >41.6 <35.2 35.2-38.6 >38.6

Median body weight (kg)


68.3 71.5 74.7 55.1 58.1 62.9
25.0-29.9 Frame index <38.6 38.6-41.8 >41.8 <35.7 35.7-38.7 >38.7

Median body weight (kg

71.8 75.9 82.2 55.6 58.6 68.7
30.0-34.9 Frame index <38.6 38.6-42.1 >42.1 <35.7 35.7-39.0 >39.0
Median body weight (kg) 74.6 72.8 85.4 57.6 60.7 72.7
35.0-39.9 Frame index <39.1 39.1-42.4 >42.4 <36.2 36.2-39.8 >39.8
Median body weight (kg)


80.4 84.1 59.5 61.8 76.7
40.0-44.9 Frame index <39.3 39.3-42.5 >42.5 <36.7 36.7-40.2 >40.2
Median body weight (kg)


79.3 84.9 59.1 62.8 77.1


Frame index <39.6 39.6-43.0 >43.0 <37.2 37.2-40.7 >40.7
Median body weight (kg)


79.8 84.0 60.3 63.4 76.8


Frame index <39.9 39.9-43.3 >43.3 <37.2 37.2-41.6 >41.6
Median body weight (kg)


78.3 83.1 60.3 64.4 77.7


Frame index <40.2 40.2-43.8 >43.8 <37.8 37.8-41.9 >41.9
Median body weight (kg) 74.8 77.9 84.5 59.9 66.3 77.6


Frame index <40.2 40.2-43.6 >43.6 <38.2 38.2-41.8 >41.8
Median body weight (kg) 73.4 76.3 80.7 60.9 64.5 76.8


Frame index <40.2 40.2-43.6 >43.6 <38.2 38.2-41.8 >41.8
Median body weight (kg) 70.3 74.5 78.9 60.2 64.9 74.5


Frame index <40.2 40.2-43.6 >43.6 <38.2 38.2-41.8 >41.8

Median body weight (kg)

70.1 72.6 76.7   60.2 62.9 74.5

Obtained from the NHANES Data as reported by Frishanco.19 Frame Index has no units and is calculated as follows: [Elbow Breadth (mm)/Height (cm)] _ 100.

Reprinted with permission.12

Table II-4. Fraction of Weight and BSA Corresponding to Amputated Limbs

Body Part Amputated
% Loss in Weight14
% of BSA to Subtract16
Leg below knee 6.5 10.0
Leg above knee
Leg at hip
Arm at elbow
Arm at shoulder



Appendix F: Detailed Rationale for Guideline 12


Assessment of Nutritional Status (Opinion)

Nutritional status of adult PD patients should be assessed on an ongoing basis in association with Kt/Vurea and CCr measurements using the Protein equivalent of Nitrogen Appearance (PNA) and Subjective Global Assessment (SGA). For pediatric PD patients, nutritional status should be assessed using the PNA and other standard nutritional assessments (see Guideline 14 of the Clinical Practice Guidelines for Peritoneal Dialysis Adequacy and the Clinical Practice Guidelines for Nutrition in Chronic Renal Failure).

Rationale Although nutritional status depends on many nondialysis-related factors, appetite suppression, nausea, and vomiting are major clinical features of inadequate dialysis. Therefore, nutritional status is also an important measure of PD adequacy. Of the available measures of nutrition, PNA is recommended because it provides an estimate of protein catabolic rate (PCR) and other protein losses. The SGA is recommended because it is a clinical assessment of patient nutritional status and is strongly associated with patient survival. Both measures are discussed in detail below.

Protein Equivalent of Nitrogen Appearance (PNA). PNA is a useful tool for monitoring the absolute level of changes in dietary protein intake. Nitrogen intake is almost entirely (95%) in the form of protein. Therefore, total nitrogen excretion in stable humans multiplied by 6.25 (there are approximately 6.25 grams of protein per gram of nitrogen) should be a good estimate of protein intake.1 This relationship does not hold true for: individuals in a state of catabolism where body cell mass may be contributing to nitrogen excretion; conditions of anabolism where the opposite occurs; and inconsistencies in absolute or time-averaged blood concentrations of BUN or creatinine.

In normal humans and in dialysis patients in nitrogen balance who have no direct protein losses in urine, dialysate or feces, the total daily excretion of nitrogen in urine, dialysate, feces, breath, and skin losses is in the form of low molecular weight nitrogenous metabolites (such as urea, creatinine, urate, amino acids, ammonia, and peptides).1

The excretion of nitrogen as low molecular weight metabolites multiplied by 6.25 approximates the amount of nitrogen in ingested protein. This calculation has been termed the protein catabolic rate (PCR).1,2 PCR actually represents the net amount of protein catabolism exceeding protein synthesis required to generate an amount of nitrogen equal to that excreted. The nitrogen in ingested protein enters the body nitrogen pools; nitrogen excreted in urine, feces, breath, skin, and dialysate represents the metabolism of a variety of body substances in these pools, such as creatine and purine, in addition to body proteins. Thus, although PCR is a reasonable estimate of protein intake, not all excreted nitrogen comes directly from protein. The protein catabolic equivalent of nitrogen excretion is actually a net catabolic equivalent, rather than an absolute. It relates directly to the contribution of protein catabolism to uremic toxicity.

In patients on hemodialysis, nitrogen balance studies have been performed to estimate total nitrogen output or appearance (skin losses were estimated and breath losses were ignored).2 The relationship of urea nitrogen appearance to total nitrogen output was assumed to be fixed and a formula was developed, known as the Borah equation, to calculate the PCR directly from urea nitrogen appearance2:

PCR (g/d) = 6.49*UNA + 0.294*V (1)

where UNA represents the net production or appearance of urea nitrogen in body fluids (any increase in body fluid nitrogen concentration times body water volume) and all measurable outputs in g/d; V is the volume of distribution of urea in liters. In hemodialysis patients with no direct protein losses in dialysate or urine, this PCR also represents an estimate of dietary protein intake and is the protein equivalent of total nitrogen appearance (PNA).

In dialysis patients with substantial urinary or dialytic protein losses (>0.1 g/kg), the direct protein losses must be added to the PCR to yield the true PNA as an estimate of dietary protein intake.3,4 Thus, in PD:

PNA = PCR + protein losses (2)

Nitrogen balance studies have also been performed in PD patients; the measured total nitrogen output (appearance) included estimates of skin and fecal nitrogen losses plus measurement of all nitrogen (including protein nitrogen) in dialysate and urine.4,5 The average daily dialysate protein loss in the CAPD patients was 7.3 grams. Urine protein losses were <1 g/24 hours. A formula for the calculation of PNA from UNA was developed:

PNA (g/d) = 10.76*(0.69*UNA + 1.46) (3)

This calculation incorporates the average dialysate protein loss of 7.3 g/d.

Calculations of PNA in PD patients with Equations 2 and 3 have been shown to yield nearly identical results.3-5 Also, subtracting protein losses in dialysate from Equation 3 yields values nearly identical to the PCR calculated by Equation 1, as developed in HD patients.4

If daily peritoneal dialysate protein losses exceed 15 grams, PNA calculated from Equation 2 will exceed PNA calculated from Equation 3 by approximately 0.1 g/kg standard body weight.5 High transporters lose more protein into effluent dialysate than other PD patients.7,8 Therefore, in high transport patients it is best to measure protein losses in dialysate directly, and if dialysate protein losses exceed 15 g/d (found in <10% of peritonitis-free patients), calculate PNA from Equation 2.5 For patients who lose large amounts of protein from any nonperitoneal source (eg, nephrotic syndrome), Equation 2 should be used.

Equations 2 and 3 have been validated with nitrogen balance studies only in CAPD and not in other therapies such as NIPD.3-5 However, since CAPD and HD patients have similar PCR values (PNA - protein losses) at any given UNA, the intermittent nature of NIPD would seem unlikely to alter the relationships. Furthermore, the daily protein losses on NIPD are similar to those of CAPD.9 As more patients and different operating variables were studied, the accuracy of formulae to predict the nPNA matured. The Bergstrom method is to obtain the nPNA surrogate for dietary protein intake by:

PNA (g/24 hours) = 15.1 + (6.95 × urea nitrogen appearance in g/24h) + dialystate and urine protein in g/24 hours6

In the absence of direct measurement of urinary and dialystate protein losses, this less accurate formula may be used:

PNA (g/24 hours) = 20.1 + (7.50 × urea nitrogen appearance in g/24 hr)

When protein losses are high, this second formula should not be used. Both formulae will require normalization to body mass in kg. These Bergstrom formulae were preferred in a small study from Italy.10

In summary, the most accurate determination of PNA in patients undergoing PD uses Equation 2, but this requires measurement of UNA and dialysate protein losses. Equation 3 is a suitable substitute and requires only measurement of UNA. However, if dialysate protein exceeds 15 g/d (many high transporters may fall into this category) and in all pediatric patients, Equation 2 is preferred.

Methods of normalizing PNA are still under debate. The Work Group recommends normalization by standard weight, which has been applied extensively. Standard weight is equal to V/0.58.11 PNA normalized by either standard weight or actual weight tends to be high in malnourished, underweight PD subjects.12 Normalization of PNA to fat-free, edema-free body mass provides appropriately low nPNA values in underweight individuals.13 The Work Group recommends that fat-free, edema-free body mass, estimated from creatinine kinetics (see Section II: Measures of PD Dose) should be used, in addition to standard weight, to normalize PNA in underweight PD subjects, defined by Table II-3 in Appendix E. Normalizing is important for patient-to-patient comparisons and to follow PNA measurements serially in an individual patient whose weight may change. If weight is stable, normalization is less important in serial measurements for an individual patient. See Guideline 14 for PNA discussion on pediatric patients.

Subjective Global Assessment (SGA). The SGA is a valid estimate of nutritional status for patients treated with PD.13 Furthermore, it is associated with the probability of patient survival.14 The SGA was developed as a clinical estimate of pre-operative nutritional status.15 For two physicians, the interobserver agreement was 72% greater than would have been predicted by chance alone. Validity was based on correlations with three measures of postoperative hospital morbidity (incidence of infection, use of antibiotics and length of stay).15 A detailed description of the SGA was provided by Detsky in 1987.16

The SGA was originally developed as a clinical assessment of preoperative nutritional status for patients prior to gastrointestinal surgery.15,16 When applied to CAPD patients,13 validity testing reduced the number of items to four (weight loss, anorexia, loss of subcutaneous tissue, and muscle mass). To increase the ability of the SGA to detect a change in nutritional status, the scoring scale was increased from a 3-point to a 7-point scale. During the development phase, the SGA was determined by physicians, research nurses, and nurse clinicians16 but was determined by dialysis nurses and dietitians when used in the CANUSA study.14

The SGA, as modified for use in CAPD patients,13 uses a 7-point scale14 which any healthcare professional can apply following a short training period.

The four items used to assess nutritional status in CAPD patients are: weight change, anorexia, subcutaneous tissue, and muscle mass.

Weight change is addressed by the question, "What was the patient’s weight change over the past 6 months?" Ideally, this should be documented by the actual weights, but historical information from the patient is acceptable. A loss of >10% is severe and 5% to 10% is moderate, while 5% is mild. This is rated subjectively on a scale from 1 to 7, where 1 or 2 is severe malnutrition, 3 to 5 is moderate to mild malnutrition, and 6 or 7 is mild malnutrition to normal nutritional status. If the weight change was intentional, the weight loss would be given less subjective weight while edema might obscure greater weight loss.

Anorexia is addressed by the question, "Has the patient’s dietary intake changed?"

Supplemental questions determine whether a decrease in dietary intake is by prescription or due to decreased appetite. Nausea and vomiting are adverse factors for this item. Again, the interviewer will rate intake on the 7-point scale with higher scores indicative of better dietary intake, better appetite, and the absence of nausea and vomiting.

Subcutaneous tissue (fat and muscle wasting) can be examined in many areas. A very detailed and illustrated brochure is available from Baxter Healthcare (publication #BRU-008-312-2000). Although the history-taking format is more detailed than required, the description of how to determine muscle wasting and subcutaneous tissue is excellent.

Subcutaneous fat can be assessed by examining the fat pads directly below the eyes and by gently pinching the skin above the triceps and biceps. The fat pads should appear as a slight bulge in a normally nourished person but are "hollow" in a malnourished person. When the skin above the triceps and biceps is gently pinched, the thickness of the fold between the examiner’s fingers is indicative of the nutritional status. The examiner then scores the observations on a 7-point scale.

Muscle mass and wasting can be assessed by examining the temporalis muscle, the prominence of the clavicles, the contour of the shoulders (rounded indicates well-nourished; squared indicates malnutrition), visibility of the scapula, the visibility of the ribs, and interosseous muscle mass between the thumb and forefinger, and the quadriceps muscle mass. These are scored on a 7-point scale.

The four item scores are then aggregated into a global score. The global score is not a simple arithmetic average of the four items. The examiner can apply different weights to the items. For example, if the physical examination items clearly indicate severe malnutrition, but the patient indicates only a moderate decrease in weight and a good appetite, the examiner might weight the physical examination items higher than the historical items.

In 23 CAPD patients four items were statistically associated with the SGA: weight loss, anorexia, loss of subcutaneous tissue, and loss of muscle mass (muscle wasting).13 Evidence for validity was provided by the correlations with serum albumin concentration, bioelectrical impedance, anthropometric measurements and normalized protein catabolic rate.

Using the SGA as originally described,16 59% of prevalent CAPD patients were well-nourished.17 Mild and severe malnutrition was reported in 33% and 8% of the patients, respectively. In 263 hemodialysis patients and 224 CAPD patients18 the SGA covaried with low visceral (ie, serum) protein concentrations, midarm muscle circumference (somatic protein mass), and body fat stores.

In the CANUSA study of peritoneal dialysis,14 weight loss, anorexia, loss of subcutaneous tissue, and loss of muscle mass (muscle wasting), as identified above as being statistically associated with the SGA,13 were used to generate the SGA for the CANUSA study. To make the scale more discriminative, the 3-point scale was expanded to a 7-point scale, with 1 and 2 corresponding to severe malnutrition, 3 to 5 corresponding to mild to moderate malnutrition, and 6 to 7 corresponding to mild malnutrition to normal nutritional status. In a multivariate analysis, a higher SGA was associated with a lower relative risk of death. A one unit increase on the 7-point scale was associated with a 25% decline in the relative risk of death (relative risk, 0.75). During the first 6 months of dialysis, the mean SGA increased 0.72 units. There was a statistically significant correlation between the increment in adequacy due to the addition of peritoneal clearance (Kt/Vurea and CCr) to RRF. Over the next 12 months, there was a small decrease in SGA and this correlated with loss of RRF estimated by CCr, but not with Kt/Vurea.19


1. Kopple JD, Jones MR, Keshaviah PR, Bergstrom J, Lindsay RM, Moran J, Nolph KD, Teehan BP: A proposed glossary for dialysis kinetics. Am J Kidney Dis 26:963-981, 1995

2. Borah MF, Schoenfeld PY, Gotch FA, Sargent JA, Wolfson M, Humphreys MH: Nitrogen balance during intermittent dialysis therapy of uremia. Kidney Int 14:491-500, 1978

3. Keshaviah PR, Nolph KD: Protein catabolic rate calculations in CAPD patients. ASAIO Trans 37:M400-M402, 1991

4. Randerson DH, Chapman GV, Farrell PC: Amino acid and dietary status in CAPD patients, in Atkins RC, Farrell PC, Thompson N (eds): Peritoneal Dialysis. Edinburgh, UK, Churchill Livingston, 1981, pp 179-191

5. Usha K, Moore H, Nolph KD: Protein catabolic rate in CAPD patients: Comparison of different techniques. Adv Perit Dial 12:284-287, 1996

6. Bergstrom J, Heimberger O, Lindholm B: Calculation of protein equivalent of total nitrogen appearance from urea appearance: which formulas should be used? Perit Dialysis Int 18:467-473, 1998

7. Malhotra D, Tzamaloukas AH, Murata GH, Fox L, Goldman RS, Avasthi PS: Serum albumin in continuous peritoneal dialysis: Its predictors and relationship to urea clearance. Kidney Int 50:1501-1507, 1996

8. Nolph KD, Moore HL, Prowant B, Twardowski ZJ, Khanna R, Gamboa S, Keshaviah P: Continuous ambulatory peritoneal dialysis with a high flux membrane. ASAIO J 904-909, 1993

9. Kathuria P, Goel S, Moore HL, Khanna R, Twardowski ZJ, Nolph KD: Protein losses during CAPD. Perit Dial Int 16:S9A, 1996 (abstr)

10. Mandolfo S, Zucchi A, D’Oro LC, Corradi B, Imbasciati E: Protein nitrogen appearance in CAPD patients: What is the best formula? Nephrol Dial Transplant 11:1592-1596, 1996

11. Nolph KD, Moore HL, Prowant B, Meyer M, Twardowski ZJ, Khanna R, Ponferrada L, Keshaviah P: Cross sectional assessment of weekly urea and creatinine clearances and indices of nutrition in continuous ambulatory peritoneal dialysis patients. Perit Dial Int 13:178-183, 1993

12. Harty JC, Boulton H, Curwell J, Heelis N, Uttley L, Venning MC, Gokal R: The normalized protein catabolic rate is a flawed marker of nutrition in CAPD patients. Kidney Int 45:103-109, 1994

13. Enia G, Sicuso C, Alati G, Zoccali C: Subjective global assessment of nutrition in dialysis patients. Nephrol Dial Transplant 8:1094-1098, 1993

14. Churchill DN, Taylor DW, Keshaviah PR: Adequacy of dialysis and nutrition in continuous peritoneal dialysis: Association with clinical outcomes. J Am Soc Nephrol 7:198-207, 1996

15. Baker JP, Detsky AS, Wesson DE, Wolman SL, Stewart S, Whitewell J, Langer B, Jeejeebhoy K: Nutritional assessment. A comparison of clinical judgement and objective measurements. N Engl J Med 306:969-972, 1982 - No Abstract Available

16. Detsky AS, McLaughlin JR, Baker JP, Johnston N, Whittaker S, Mendelson RA, Jeejeebhoy KN: What is subjective global assessment of nutritional status? J Parenter Enteral Nutr 11:8-13, 1987

17. Young GA, Kopple JD, Lindholm B, Vonesh EF, DeVecchi A, Scalamogna A, Castelnovo C, Oreopoulos DG, Anderson GH, Bergstrom J, Dichiro J, Gentile D, Nissenson A, Sakhrani L, Brownjohn A, Nolph KD, Prowant BF, Algrim CE, Martis L, Serkes KD: Nutritional assessment of continuous ambulatory peritoneal dialysis patients: An international study. Am J Kidney Dis 17:462-471, 1991

18. Cianciaruso B, Brunori G, Kopple JD, Traverso G, Panarello G, Enia G, Strippoli P, De Vecchi A, Querques M, Viglino G, Vonesh E, Maiorca R: Cross-sectional comparison of malnutrition in continuous ambulatory peritoneal dialysis and hemodialysis patients. Am J Kidney Dis 26:475-486, 1995

19. McCusker FM, Teehan BP, Thorpe K, Keshaviah P, Churchill D: How much peritoneal dialysis is required for the maintenance of a good nutritional state? Kidney Int 50:S56-S61, 1996

Appendix G: Detailed Rationale for Guideline 15


Weekly Dose of CAPD (Evidence)

For CAPD, the delivered PD dose should be a total Kt/Vurea of at least 2.0 per week and a total creatinine clearance (CCr) of at least 60 L/wk/1.73 m2 for high and high-average transporters, and 50 L/wk/1.73 m2 in low and low-average transporters.

Rationale The evidence supporting this guideline is derived from theoretical constructs and cohort studies which use either univariate or multivariate statistical analyses.

The original description of CAPD1 suggested that an anephric 70 kg patient with a total body water of 42 L would remain in nitrogen balance with a daily dialysis prescription of 10 L given as five 2-L exchanges. Full equilibration of urea between plasma and dialysate and 2 L/day of net ultrafiltration were assumed. This would produce a daily urea clearance of 12 L or a weekly urea clearance of 84 L. For a patient with a total body water of 42 L, this corresponds to a weekly Kt/Vurea of 2.0. Others, using the concept of the Dialysis Index, suggested that a similar patient would require 13.5 L daily of equilibrated drained dialysate to maintain nitrogen balance, and this would produce a weekly Kt/Vurea of 2.25.2 The difference between these two projections is due to the higher target protein intake used in the latter calculation. Using the peak urea concentration hypothesis, a weekly Kt/Vurea of 2.0 is equivalent to a single pool hemodialysis Kt/Vurea of 1.3 for patients receiving thrice weekly dialysis.3 These theoretical constructs suggest that a weekly Kt/Vurea of 2.0 to 2.25 would be appropriate.

Validation of these theoretical constructs requires clinical study. A series of cohort studies addressed this issue.4-8 Initially no relationship was found between urea clearance and patient survival,6 but a reanalysis, using an anthropometric9 estimate for total body water, found that patients with a weekly Kt/Vurea <1.5 had an increased risk of death compared to patients with a weekly Kt/Vurea 1.5. In another study10 a mean weekly Kt/Vurea >1.89 was associated with a decreased risk of death compared to patients with less dialysis, while yet another4 reported that patients surviving for a 12-month follow-up had a mean weekly Kt/Vurea of 2.0 compared to a mean of 1.7 among those who did not survive for 12 months. A Belgian group reported that 16 patients surviving 5 years on CAPD had a mean weekly Kt/Vurea of 2.0.5 These studies all used univariate analysis and therefore did not simultaneously evaluate the association between other important variables (eg, age, diabetes, cardiovascular disease) and patient survival.

Several studies have used multivariate statistical analysis to evaluate the association between adequacy of PD and survival while controlling for other variables.7,8,11,12 In one such study a lower serum albumin concentration, increased age, greater time on dialysis, and lower weekly Kt/Vurea were associated with a decreased probability of patient survival.11 A French group reported that patients with a weekly Kt/Vurea >1.7 and a weekly CCr of >50 L/1.73 m2 at initiation of dialysis had better survival than those with lower values at initiation.7 However, these investigators did not evaluate the effect of changes in adequacy over time due to loss of RKF, nor did they attempt to evaluate any association of higher weekly Kt/Vurea or CCr with survival. An Italian group evaluated the association between estimates of adequacy and patient survival in a cohort of 68 prevalent continuous PD patients followed over 3 years.8 A mean weekly Kt/Vurea of 1.96 was associated with better survival than lower values. No further benefit was observed with a Kt/Vurea higher than 1.96. Among these patients, a weekly Kt/Vurea of 1.96 corresponded to a weekly CCr of 58 L/1.73 m2.

The Canada-USA (CANUSA) study evaluated the association between adequacy of PD and patient survival, technique survival, and hospitalization among 680 incident patients (new to starting PD) treated with continuous PD.12 A decrease of 0.1 in weekly Kt/Vurea was associated with a 5% increase in the relative risk of death, and a decrease of 5 L/1.73 m2/wk in CCr was associated with a 7% increase in the risk of death. The risk of technique failure increased with decreased creatinine clearance, but was not associated with Kt/Vurea. Hospitalization increased with decreased CCr. Using data derived from the multivariate analysis, the predicted 2-year survival associated with a constant weekly Kt/Vurea of 2.1 was 78%. The corresponding weekly CCr was 70 L/1.73 m2.

Thus, there is both a theoretical rationale and convincing evidence supporting an association between greater clearance of urea and creatinine and better patient survival. There is also evidence supporting an association between greater CCr to longer technique survival and less hospitalization. In summary, theoretical constructs1-3 suggest that the minimum weekly Kt/Vurea should be 2.0. Cohort studies using univariate statistical analysis support this "target."4-8,11,12 The CANUSA study predicts, among North American patients, a 78% 2-year survival with a weekly Kt/Vurea of 2.1.12

There are no theoretical data to support a specific CCr target. The CCr which corresponds to a weekly Kt/Vurea of 2.1 in the CANUSA study was 70 L/1.73 m2/wk. The Italian group found that a weekly Kt/Vurea of 1.96 corresponded to a weekly CCr of 58 L. The CANUSA study involved incident patients with significant RKF, while the Italian study evaluated prevalent patients with much less RKF.8 The target of 60 L/1.73 m2/wk was selected by the Work Group because it is more relevant to patients with diminished renal function.

Even after controlling for delivered dose, low and low-average transporters have better patient and technique survival outcomes than do high and high-average transporters.13 In the absence of adequate residual renal function, low and low-average transporters may not be able to achieve a CCr of 60 L/wk/1.73 m2 on any reasonable dialysis prescription. However, because urea clearance is less affected than creatinine clearance by transport status, low and low-average transporters can achieve a weekly Kt/V of 2.0. Therefore, it seems reasonable to lower the CCr target in low and low-average transporters without jeopardizing the outcomes. These patients must be observed closely for evidence of inadequate dialysis.

There are few data to address the issue of adequate compared to optimal dialysis. The latter is defined in part as the dialysis dose above which the incremental clinical benefit is not justified by the social cost to the patient or the financial cost to society. Whether or not increased weekly Kt/Vurea greater than 2.0 will be associated with improved clinical outcomes requires further study.

The relative importance of RKF compared to peritoneal clearance and the relative importance of urea compared to CCr are important and interrelated issues. The convention has been to consider RKF and peritoneal clearance to be equivalent and therefore additive. Some believe that renal clearance is more important, but in the absence of data establishing the magnitude of that difference, the assumption of equivalence was adopted by the Work Group. CCr appeared more important than urea clearance in the CANUSA study.12 The former was associated with patient survival, technique survival, and hospitalization, while the latter was associated only with patient survival. One potential explanation for this finding is that CCr is more strongly associated with better RKF than was Kt/Vurea. This explanation is based on the assumption that RKF is better than peritoneal clearance, an opinion not yet supported by evidence.

The equivalence of peritoneal and residual kidney clearance is controversial. Current data suggest inconsistent conclusions. There is a strong suggestion that protein metabolism is similar in patients with progressive chronic kidney disease.14 Peritoneal clearance has been shown to predict survival.15,16 However, a large retrospective analysis suggested that peritoneal clearance was not predictive of survival, while residual kidney clearance was.17 Thus, the Work Group has adopted the position that until more definitive data are available to direct us, the simplest solution is to continue to equate residual kidney and peritoneal clearance. To that end, the preservation of kidney clearance is paramount and strategies to achieve this have recently been described.18

Until evidence to the contrary is available, the Work Group recommends that kidney and peritoneal clearances be considered equivalent. If there is discordance between achieving the target Kt/Vurea and CCr, the Kt/Vurea should be the immediate determinant of adequacy since it reflects protein catabolism. However, the reason for the discrepancy should be sought and the patient monitored closely for clinical signs of underdialysis.

A special case is the underweight patient, defined in Table II-3, Appendix E. Successful efforts to restore weight to a normal level in such a patient will result in a rising V, and consequently in a proportionally declining Kprt/Vurea. To provide a weekly Kprt/Vurea of 2.0 at the final increased weight, the weekly target Kprt/Vurea provided during the malnourished state must be greater than 2.0. The Work Group recommends that the target Kprt/Vurea should be raised in a malnourished CAPD patient to the level that would provide a weekly Kprt/Vurea of 2.0 for that patient if he or she was at normal weight. That level is calculated by multiplying the target of 2.0 for CAPD times the ratio of Vdesired/Vactual. This is described in detail in Appendix E: Detailed Rationale for Guideline 9, and discussed in Guideline 17: PD Dose in Subpopulations. The same upward target adjustment would be made in CCr. The target CCr should be increased by a factor of BSAdesired/BSAactual.

Clinical judgment suggests that the target doses of PD for children should meet or exceed the adult standards. However, there are currently no definitive outcome data in pediatrics to suggest that any measure of dialysis adequacy is predictive of well-being, morbidity, or mortality. There are limited data regarding the real protein needs of children, especially young children, on dialysis. It is the opinion of the Work Group that the nutritional requirements per kilogram of body weight are higher in children than in adults. Therefore, PD doses in children, and especially small infants who have very high protein intakes, may have to be higher than PD doses in adults.


1. Popovich RP, Moncrief JW: Kinetic modeling of peritoneal transport. Contrib Nephrol 17:59-72, 1979 - No Abstract Available

2. Teehan BP, Schleifer CR, Sigler MH, Gilgore GS: A quantitative approach to the CAPD prescription. Perit Dial Bull 5:152-156, 1985

3. Keshaviah PR, Nolph KD, Prowant B, Moore H, Ponferrada L, Van Stone J, Twardowski ZJ, Khanna R: Defining adequacy of CAPD with urea kinetics. Adv Perit Dial 6:173-177, 1990

4. De Alvaro F, Bajo MA, Alvarez-Ude F, Vigil A, Molina A, Coronel F, Selgas R: Adequacy of peritoneal dialysis: Does Kt/V have the same predictive value as in HD? A multicenter study. Adv Perit Dial 8:93-97, 1992

5. Lameire NH, Vanholder R, Veyt D, Lambert M, Ringoir S: A longitudinal, five year survey of urea kinetic parameters in CAPD patients. Kidney Int 42:426-432, 1992

6. Blake PG, Sombolos K, Abraham G, Weissgarten J, Pemberton R, Chu GL, Oreopoulos DG: Lack of correlation between urea kinetic indices and clinical outcomes in CAPD patients. Kidney Int 39:700-706, 1991

7. Genestier S, Hedelin G, Schaffer P, Faller B: Prognostic factors in CAPD patients: A retrospective study of a 10-year period. Nephrol Dial Transplant 10:1905-1911, 1995

8. Maiorca R, Brunori G, Zubani R, Cancarini GC, Manili L, Camerini C, Movilli E, Pola A, d’Avolio G, Gelatti U: Predictive value of dialysis adequacy and nutritional indices for mortality and morbidity in CAPD and HD patients. A longitudinal study. Nephrol Dial Transplant 10:2295-2305, 1995

9. Blake PG, Balaskas E, Blake R, Oreopoulos DG: Urea kinetics has limited relevance in assessing adequacy of dialysis in CAPD. Adv Perit Dial 8:65-70, 1992

10. Teehan BP, Schleifer CR, Brown J: Urea kinetic modeling is an appropriate assessment of adequacy. Semin Dial 5:189-192, 1992

11. Teehan BP, Schleifer CR, Brown JM, Sigler MH, Raimondo J: Urea kinetic analysis and clinical outcome on CAPD. A five year longitudinal study. Adv Perit Dial 6:181-185, 1990

12. Churchill DN, Taylor DW, Keshaviah PR: Adequacy of dialysis and nutrition in continuous peritoneal dialysis: Association with clinical outcomes. J Am Soc Nephrol 7:198-207, 1996

13. Churchill D, Thorpe K, Nolph K, Keshoviah P, Oreopoulos P, Page D: Increased peritoneal membrane transport is associated with decreased patient and technique survival for continuous peritoneal dialysis patients. J Am Soc Nephrol 9:1285-1292, 1998

14. Mehrotra R, Saran R, Moore H, Prowant B, Khanna R, Twardowski Z, Nolph K: Toward targets for initiation of chronic dialysis. Perit Dial Int 17:497-508, 1997

15. Jager KI, Merkus M, Dekker F, Boeschoten E, Tijssen J, Stevens P, Bos WJ, Krediet R: Mortality and technique failure in patients starting chronic peritoneal dialysis: Results of the Netherlands Cooperative Study on the adequacy of dialysis. Kidney Int 55:1476-1485, 1999 - Full Text

16. Davies S, Phillips L, Russel G: Peritoneal solute transport predicts survival on CAPD independently of residual renal function. Nephrol Dial Transplant 13;962-968, 1998

17. Diaz-Buxo J, Lowrie E, Lew N, Zhang H, Zhu X, Lazarus JM: Associates of mortality among peritoneal dialysis patients with special reference to preitoneal transport rates and solute clearance. Am J Kidney Dis 33:523-534, 1999

18. Lamiere N, Van Biesen W: The impact of residual renal function on the adequacy of peritoneal dialysis. Perit Dial Int 17:S102-S110, 1997 (suppl 2)

Appendix H: Detailed Rationale for Guideline 19


Identify and Correct Patient-Related Failure to Achieve Prescribed PD Dose (Opinion)

Potential patient-related causes of failure to achieve prescribed peritoneal dialysis dose should be investigated and corrected. These include:

• Failure to comply with the prescription.

• Lack of understanding of the importance of adherence to the full prescription.

• Sampling and collection errors.

Rationale It is the opinion of the Work Group that to increase the likelihood of achieving a prescribed dose of PD, it is necessary to elucidate the patient-related causes of failure to achieve a prescribed dose. Selection of inappropriate candidates for PD may result in failure to achieve a prescribed dose due to medical, technical, and/or psycho-social reasons. The issue of medically appropriate patient selection is dealt with at length in Section VIII: Suitable Patients for PD. In addition to the medical reasons for selecting patients for PD or HD discussed in Section VIII, patient compliance is of paramount importance and should be explored.

Failure to Comply With the Prescription. Patients may decrease the delivered dose of PD in several ways. Some of the ways are listed below:

• Skipping exchanges.

• Shortened exchange times.

• Dialysate dumping: too much flushing resulting in too little fill.1

• Delayed dumping: This is achieved by partially draining before the dwell is completed.

• Reduction of total cycler time.

• Unscheduled dry days on CCPD.

A validated method to measure patient compliance is not currently available. Methods proposed for evaluating compliance include monitoring for variations in creatinine output in dialysate and urine2 as detailed in Guideline 7: PD Dose Troubleshooting. Evidence for this recommendation is currently not available. The recommendation represents, therefore, the opinion of the Work Group members.

In the absence of a validated method to measure patient compliance, its prevalence in the PD population is not known. Preliminary data from the USRDS DMMS Wave II project show that 487 CAPD patients self-report full compliance with 82.8% of their exchanges.3 One exchange/week is missed by 11.5% of patients and 2 to 3 exchanges/week are missed by 4.5% of patients, all self-reported. Other estimates vary between 5% and 38%.4 Thus, noncompliance is a major cause of a delivered PD dose being less than the desired dose and is potentially preventable.

Lack of Understanding of the Importance of Adherence to the Full Prescription. Medical literature about conditions associated with noncompliance in PD is inadequate. The Work Group reviewed published information on compliance in hemodialysis5-8 and in drug treatment for chronic illness.9-11 The Work Group believes that some of the conclusions in this literature may be applicable to PD. An important conclusion of the studies on drug compliance is that lack of education regarding the importance of adherence to the full prescription partially contributes to compliance failure. Compliance with the drug prescription improves when the patient is convinced that the diagnosis is accurate, the reasons for the prescribed treatment are correct, and the prescribed treatment is beneficial.11 Some contend that patients on dialysis are more likely to follow the prescribed treatment if they can be convinced that adherence to the prescription is in their own interest.8 By not understanding the significance, importance, value, or relevance of the collections or the exchanges, noncompliance could occur without patient concern. Therefore, proper education about the treatment may increase compliance in many PD patients. Patients should be educated that dialysis prescription may change over time (different modality and/or increase in the number or volume of exchanges) due to loss of residual renal function (see Guideline 6: Assessing Residual Renal Function). The method of education should emphasize the expected positive results (improved survival, well-being) of adherence to the PD prescription, rather than the negative outcomes (morbidity, mortality) of nonadherence, to prevent the development of excessive anxiety, which has adverse effects on compliance.

Patient education should be continuous throughout the course of PD. Patients should be told the results of the repeated clearance measurements and should be aware of the target values for Kprt/V and CCr and of the clinical significance of these clearances. Prevention of noncompliance should include monitoring the patient’s psychological status. In studies on compliance to drugs, certain psychiatric conditions, such as hostility toward authority, depression and memory impairment, financial problems, impaired mobility, and language or ethnic barriers, have been associated with poor compliance.11 In addition, complexity of the prescription9 and chronicity of the treatment11 increased noncompliance. In the case of prolonged treatment, repetition of the teaching at 6-month intervals improved compliance.11 In studies on compliance in HD, male gender12 and young age13 were predictors of poor compliance with different aspects of HD prescription. Preliminary information suggests that a general negative attitude of the patients predicts noncompliance in PD.14 Finally, drug compliance improves with better education of the providers about compliance issues.10 The Work Group thinks that all of these issues are relevant to PD. The psychological profile which is predictive of noncompliance and the best method of characterizing this profile should be a subject for research in the future. For the present, the Work Group’s opinion is that monitoring of patients’ psychological status should be aimed at detecting conditions associated with increased risk of noncompliance, and particularly at detecting a negative patient attitude towards PD. Teaching patients about their PD prescription should be repeated at intervals of 6 months or less.

Sampling and Collection Errors. Sampling and collection errors committed by patients during the clearance study preclude accurate measurement of clearance. Such errors include:

• Batch method: the patient may not recognize the importance of accidentally spilled dialysate.

• Aliquot method: Inaccurate weighing of drain volumes which may be the result of inaccurate scales or misreading. Disproportionate filling of the syringe with dialysate.

• Errors in weighing bags for variable fill volumes: for example, when (due to cost issues) 3-L bags are being used and only 2.5 L are exchanged.

• Incomplete urine collections for the RRF determination.

Many of these errors can be prevented by careful patient instruction about the details and significance of the clearance procedure.


1. Caruana RJ, Smith KL, Hess CP, Perez JC, Cheek PL: Dialysate dumping: A novel cause of inadequate dialysis in continuous ambulatory peritoneal dialysis (CAPD) patients. Perit Dial Int 9:319-320, 1989

2. Keen M, Lipps B, Gotch F: The measured creatinine generation rate in CAPD suggests only 78% of prescribed dialysis is delivered. Adv Perit Dial 9:73-75, 1993

3. US Renal Data Systems: the USRDS Dialysis Morbidity and Mortality Study (Wave 1), in National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases (ed): US Renal Data Systems 1997 Annual Data Report. Chapter 4. Bethesda, MD, 1997, pp 49-67

4. Amici G, Viglino G, Virga G, Gandolfo C, Da Rin G, Bocci C, Cavalli PL: Compliance study in peritoneal dialysis using PD Adequest software. Perit Dial Int 16:S176-S178, 1996

5. Wolcott DL, Maida CA, Diamond R, Nissenson AR: Treatment of compliance in end-stage renal disease. Am J Nephrol 6:329-338, 1986

6. King K: Noncompliance in the chronic dialysis population. Dial Transplant 20:67-68, 1991

7. Rocco MV, Burkart J: Prevalence of missed treatments and early sign-offs in hemodialysis patients. J Am Soc Nephrol 4:1178-1183, 1993

8. Lundin AP: Causes of noncompliance in dialysis patients. Dial Transplant 24:174-176, 1995

9. Porter AM: Drug defaulting in a dialysis patients. BMJ 1:218-222, 1969

10. Innui TS, Yourtee GL, Williamson JW: Improved outcomes in hypertension after physician tutorials. A controlled trial. Ann Intern Med 84:646-651, 1976

11. Anderson RJ, Kirk LM: Methods for improving patient compliance in chronic disease states. Arch Intern Med 142:1673-1675, 1982

12. Everett KD, Sletten C, Carmach C, Brantley PJ, Jones GN, McKnight LT: Predicting noncompliance to fluid restrictions in hemodialysis patients. Dial Transplant 22:614-622, 1993

13. Cummings KM, Becker MH, Kirscht JP, Levin NW: Psychosocial factors affecting adherence to medical regimens in a group of hemodialysis patients. Med Care 20:567-580, 1982

14. Bernardini J: Predicting compliance in home peritoneal dialysis (PD) patients. ASAIO J 42:102A, 1996 (abstr)




© 2001 National Kidney Foundation, Inc

web version created by cyberNephrologyTM and The Nephron Information Center