THE IMPACT OF BOWEL PREPARATION FOR COLONOSCOPY ON SERUM ELECTROLYTES: POLYETHYLENE GLYCOL-ELECTROLYTE SOLUTION VERSUS SODIUM PHOSPHATE

Abstract

Background Polyethylene glycol-electrolyte solution (PEG) and sodium phosphate (NaP) are the two representatives used for bowel cleansing prior to colonoscopy. Because of the possibility of acute phosphate nephropathy, NaP has largely been replaced by PEG. However, large-scaled controlled studies are required to demonstrate whether no significant serum electrolyte disturbances are induced by PEG.MethodsFasting serum data were collected from the adults who had regular checkups at Hanyang University Hospital in 2011. According to the use of bowel preparation solution, the subjects were grouped into PEG (n=2593), NaP (n=826), and controls (non-users, n=3839). Those with eGFR < 60 ml/min/1.73㎡ were excluded. Individual serum levels of the measured sodium, potassium, chloride, calcium and phosphate were compared between the two and three groups using one-way ANOVA and Scheffe post hoc test. P-value was adjusted by age, sex and eGFR.ResultsCompared with controls (Na, 140.3 + 1.9), PEG and NaP group (PEG, 141.0 ± 2.2; NaP, 141.3 ± 2.7) showed significantly higher (P<0.05) serum sodium levels. Hypernatremia (Na>145 mEq/L) was more frequently observed in PEG (1.6%) than controls (0.5%). Also, hyponatremia incidence of PEG group is about two to three times more than control group. In particular, high risk of hypernatremia and hyponatremia associated with aging. Compared with controls (K, 4.18 + 0.34; Ca, 9.22 + 0.35; P, 3.45 + 0.48), serum potassium, calcium and phosphorus were significantly altered in NaP (K, 3.80 + 0.38; Ca, 8.74 + 0.44; P, 6.71 + 1.79) (P<0.05), but not in PEG (K, 4.16 + 0.35; Ca, 9.24 + 0.39; P, 3.45 + 0.59) (P>0.05). However, hypokalemia (K<3.5 mEq/L), hypocalcemia (Ca<8.4 mg/dL), and hyperphosphatemia (P>4.5 mg/dL) were more frequently observed in PEG (1.3, 1.6, and 2.4%) than controls (0.7, 0.7, and 1.1%) (P<0.0001). Women had high risk of hypokalemia, hypocalcemia and hyperphosphatemia (P<0.05). High risk of hypokalemia, hypocalcemia and hyperphosphatemia were associated with creatinine decrease (P<0.05). Furthermore, high risk of hypocalcemia was associated with aging and high risk of hyperphosphatemia was associated with eGFR increase (P<0.05).ConclusionSodium concentration of PEG group is higher than control group. It seems like this difference has not clinically meaningful result. But, hypernatremia incidence of PEG group is three times more than control group, and hyponatremia incidence of PEG group is about two to three times more than control group. In particular, high risk of hypernatremia and hyponatremia were associated with aging. In our study, PEG induces no disturbances in serum potassium, calcium and phosphorus. However, hypokalemia, hypocalcemia, and hyperphosphatemia were more frequently observed in PEG than controls. Women and creatinine decrease had high risk of that electrolyte disturbance. Moreover, hypocalcemia and hyperphosphatemia were associated with aging and eGFR increase, respectively.