Table 3.

Quality assessment of four cohort studies using a modified Newcastle–Ottawa Scale.

Study	Criteria of modified Newcastle–Ottawa Scale
	Prior Criteria; Representativeness	Ascertainment of Exposure	Starting Condition Prior to Outcome	Adjustment for Confounding Factors a	Ascertainment of Outcome	Sufficiency of Follow-Up Period	Adequacy of Follow-Up
Hsiao 2010 [21]	Lack of detailed description of selection criteria	Medical records	Chyloperitoneum was not pre sent	Infection and other causes	Standardised assessment	No statement	Complete follow-up
Yang 2008 [23]	Lack of detailed description of selection criteria	Medical records	Chyloperitoneum was not present	Infection and other causes	Standardised assessment	Follow-up period ≥ 30 d	More than 90%
Topal 2006 [22]	Unclear	Medical records	Chyloperitoneum was not present	Infection	Visual observation; Turbidity of peritoneal dialysate	Not sufficient; 1 d	Complete follow-up
Yoshimoto 1998 [11]	Participant selection by researchers	Medical records	Chyloperitoneum was not present	Infection and other causes	Standardised assessment	Not sufficient; 2 d	Subsequent observation by a biochemical test was performed on 53% of the patients

^a The factors except for infection were possible causes of chyloperitoneum, e.g., cancer, lymphatic obstruction, and traumatic abdominal injury. Abbreviation: d, day(s).