Table 3.
Summary of GRADE findings: Noncalcium-based phosphate binders compared with placebo
| Outcomes | Illustrative Comparative Outcomes | Relative Effect (95% Cl) | Number of Participants (studies) | Certainty of Evidence (GRADE) | Comments | |
|---|---|---|---|---|---|---|
| Effect of non calcium-based phosphate binders | Risk with placeboa | |||||
| Serum phosphate (mg/dl) (3–36 months) | The measure of mean serum phosphate was -0.37 mg/dl lower than the placebo group | The mean serum phosphate was 4.28 mg/dl | −0.58 to -0.15 | 1280 (12) | ⊕⊕OO Due to risk of bias and inconsistency |
|
| Urinary phosphate excretion (SMD) (3–24 months) | A medium effect on lowering urinary phosphate with the use of phosphate binders (Hedges g=0.61) | The mean urinary phosphate was 130.52 standard units | −0.90 to -0.31 | 702 (8) | ⊕⊕OO Due to risk of bias and inconsistency |
Different units of urinary phosphate excretion including ratios added to inconsistency of results |
| Serum PTH (pg/dl) (3–24 months) | The measure of mean serum PTH was unchanged | The mean serum PTH was 432.5 pg/dl | −18.79 to 8.92 | 1080 (10) | ⊕⊕ΟΟ Due to risk of bias and inconsistency of results |
Some inconsistency in results regarding effect on PTH. Heterogeneity of studies (I2=70%) |
| Pulse wave velocity (m/s) (10–24 months) | The PWV was unchanged | The mean PWV was 10.4 m/s | −0.46 to 1.05 | 333 (3) | ⊕OOO Due to risk of bias, imprecision and inconsistency |
Three studies only able to be analyzed, with two sets of mean values identical (likely due to transformation of median to mean for analysis) |
| Vascular calcification (Standard units) (9–24 months) | A low effect of increasing vascular calcification was noted with the use of phosphate binders (Hedge’s g=0.41) | High background risk of vascular calcification in CKD depending on stage (30%–70% of CKD population has defined vascular calcification by stage 3–4 CKD) | 0.17 to 0.77 | 184 (3) | ⊕ΟΟΟ Due to risk of bias, imprecision and inconsistency |
Statistically significant (P<0.01)/ Not likely to be of clinical significance. Three studies only able to be analyzed. |
| Mortality (log OR) (3–24 months) | There was no significant difference in mortality rate between the two groups (log OR -0.51) | Risk with placebo 2/100 person years64 | −2.12–1.11 | 1479 (8) | ⊕OOO Due to risk of bias, imprecision and inconsistency |
Large confidence intervals leading to imprecision. Studies often not powered or long enough follow-up for consideration of mortality |
| Cardiovascular events (log OR) (3–24 mos) | There was no significant difference in cardiovascular events between the two groups (log OR 0.31) | Not estimable | −0.55–1.17 | 873 (7) | ⊕OOO Due to risk of bias, imprecision and inconsistency |
Challenge due to differing follow-up depending on studies |
a
Calculations for overall weighted means for serum and urinary phosphate, PTH and PWV– sum of weighted placebo means across studies for each variable.