Skip to main content

Table 2 Sensitivity analysis of the risk of serious adverse events associated with hyaluronic acid for osteoarthritis of the knee

From: Hyaluronic acid injection therapy for osteoarthritis of the knee: concordant efficacy and conflicting serious adverse events in two systematic reviews

 

Relative risk (RR) (95 % confidence interval (CI))

RR (95 % CI) excluding outlier study by Jubb et al. [64]

Pooled SAE results in our review

1.39 (0.78–2.47)

Original pooled SAE results in Rutjes et al. [8]

  

- All studies

1.41 (1.02–1.97)

- Sample restricted to “large, blinded trials”

1.55 (1.07–2.24)

Our approximated pooled SAE results (data imputed for studies in Rutjes et al. [8] with proprietary data)

  

- All studies

1.42 (1.01–1.99)

1.26 (0.83–1.90)

- Sample restricted to “large, blinded trials”

1.54 (1.05–2.28)

1.37 (0.83–2.26)

Approximated pooled SAE results including omitted SAEs (two cases of myocardial infarction [54, 75], four cases of severe knee swelling [54], one case of cerebral hemorrhage [26], and one case of breast cancer [75])

  

- All studies

1.37 (0.98–1.91)

1.21 (0.82–1.80)

- Sample restricted to “large, blinded trials”

1.50 (1.03–2.19)

1.32 (0.82–2.13)

Approximated pooled SAE results excluding four cancer cases [73]

  

- All studies

1.36 (0.97–1.89)

1.20 (0.81–1.77)

- Sample restricted to “large, blinded trials”

1.48 (1.02–2.16)

1.30 (0.81–2.09)

Approximated pooled SAE results including omitted non-cancer SAEs (two cases of myocardial infarction [54, 75], four cases of severe knee swelling [54], one case of cerebral hemorrhage [26]) and excluding four cancer cases [73]

  

- All studies

1.37 (0.98–1.91)

1.21 (0.81–1.80)

- Sample restricted to “large, blinded trials”

1.50 (1.03–2.18)

1.32 (0.82–2.12)

  1. This table compares the SAE results in our review, the review by Rutjes et al. [8], our approximation of their results, and the sensitivity of those approximated results to the inclusions and exclusion of individual AEs and the outlier study by Jubb et al. [64]