Table 6 Quality assessment of observational studies examining risk of nephrocalcinosis/nephrolithiasis (NC/NL) in premature infants
Study (Year) | Risk of Bias (Low, Moderate, Serious, Critical, No Information) | Comments |
|---|---|---|
Hufnagle (1982) [30] | Critical | Critical risk of bias in confounding domain: no statistical tests performed in the analysis to test association of NC and furosemide. Critical risk of bias in selection of participants into the study: All infants were exposed to furosemide and had NC; lack of control group. |
Woolfield (1988) [31] | Serious | Serious risk of bias in confounding domain: no statistical tests performed in the analysis to test association of NC and furosemide. |
Jacinto (1988) [32] | Serious | Serious risk of bias in confounding domain: Lower BW and GA associated with outcome (NC), along with exposure to furosemide. Did not control for severity of illness. |
Ezzedeen (1988) [33] | Serious | Serious risk of bias in confounding domain: no adjustment for severity of illness; small number of infants in control group. |
Short (1991) [34] | Moderate | Multivariate analyses controlling for other risk factors for NC. Dose-response relationship evaluated. |
Downing (1991) [35] | Moderate | All infants screened for the outcome had a diagnosis of chronic lung disease; high percentage of follow-up imaging obtained. |
Downing (1992) [36] | Moderate | Robust comparators; long-term follow-up. |
Stafstrom (1992) [37] | Serious | Serious risk of bias in confounding domain: no statistical tests performed in the analysis to test association of NC and furosemide. No reporting of frequency of NC in infants with post-hemorrhagic hydrocephalus not exposed to furosemide. |
Pope (1996) [38] | Moderate | Similar severity of illness in each group; long-term follow up with serial ultrasounds. Dose-response relationship evaluated. |
Saarela (1999) [39] | Moderate | Dose-response relationship evaluated. |
Schell-Feith (2000) [40] | Moderate | Large sample size. Control group without NC included. |
Narendra (2001) [41] | Moderate | Multivariate analyses controlling for other risk factors for NC. Dose-response relationship evaluated. |
Hoppe (2002) [42] | Serious | Serious risk of bias in confounding domain: Lack of control group without NC. |
Hein (2004) [43] | Moderate | Large sample size with appropriate control groups. |
Ketkeaw (2004) [44] | Moderate | Appropriate control group included. Dose-response relationship evaluated. |
Cranefield (2004) [45] | Moderate | All infants with comparable severity of illness. |
Gimpel (2010) [46] | Moderate | Multivariate analyses controlling for other risk factors for NC. Dose-response relationship evaluated. |
Chang (2011) [47] | Serious | Serious risk of bias in confounding domain: no adjustment for severity of illness. Low incidence of NC in sample. |
Lee (2014) [48] | Serious | Serious risk of bias in confounding domain: no adjustment for severity of illness. |
Mohamed (2014) [49] | Serious | Serious risk of bias in confounding domain: no adjustment for severity of illness. |