Table 4. Statistical Reporting in Bayesian analysis [n/N(%)].
| Items | All studies (n = 61) | Journal impact factor* | Year of publication | ||||
|---|---|---|---|---|---|---|---|
| Low (<5.00)(n = 26) | High (≥5.00)(n = 26) | P-value | Older studies(~2013) (n = 29) | Recent studies(2014~2015) (n = 32) | P-value | ||
| Was traditional meta-analysis conducted? (Yes) | 42/61(68.85) | 15/26(57.69) | 20/26(76.92) | 0.143 | 18/29(62.07) | 24/32(93.10) | 0.280 |
| Were summary measures reported? (Yes) | 57/61(93.44) | 24/26(92.31) | 25/26(96.15) | 0.556 | 27/29(93.10) | 30/32(93.75) | 0.920 |
| Planned methods of analysis: | |||||||
| What was the model used? | |||||||
| Fixed-effect | 7/46(15.22) | 4/20(20.00) | 2/24(8.33) | 0.267 | 6/26(23.08) | 1/20(5.00) | 0.094 |
| Random-effects | 11/46(23.91) | 4/20(20.00) | 6/24(25.00) | 0.697 | 6/26(23.08) | 5/20(25.00) | 0.881 |
| Fixed- and random-effects | 18/46(39.13) | 12/20(60.00) | 6/24(25.00) | 0.020 | 7/26(26.92) | 11/20(55.00) | 0.056 |
| Other models | 10/46(21.74) | 0/20(0.00) | 10/24(41.67) | 0.001 | 7/26(26.92) | 3/20(15.00) | 0.336 |
| Nor reported | 15/61(24.59) | 6/26(23.08) | 2/26(7.69) | 0.128 | 3/29(10.34) | 12/32(37.50) | 0.015 |
| How to present the model code? | |||||||
| Supplement | 5/21(23.81) | 2/6(33.33) | 3/15(20.00) | 0.527 | 2/14(14.29) | 3/7(42.86) | 0.157 |
| Reference | 16/21(76.19) | 4/6(66.67) | 12/15(80.00) | 0.527 | 12/14(85.71) | 4/7(57.14) | 0.157 |
| Not provided | 40/61(65.57) | 20/26(76.92) | 11/26(42.31) | 0.012 | 15/29(51.72) | 25/32(78.13) | 0.032 |
| How to assess the model fit? | |||||||
| Residual deviance | 2/24(8.33) | 1/14(7.14) | 1/10(10.00) | 0.807 | 1/9(11.11) | 1/15(6.67) | 0.709 |
| Deviance information criterion | 15/24(62.50) | 8/14(57.14) | 7/10(70.00) | 0.530 | 6/9(66.67) | 9/15(60.00) | 0.749 |
| Residual deviance+Deviance information criterion | 7/24(29.17) | 5/14(35.71) | 2/10(20.00) | 0.414 | 2/9(22.22) | 5/15(33.33) | 0.570 |
| Not reported | 37/61(60.66) | 12/26(46.15) | 16/26(61.54) | 0.270 | 20/29(68.97) | 17/32(53.13) | 0.210 |
| Was handling of multigroup trials reported? (Yes) | 5/61(8.20) | 1/26(3.85) | 3/26(11.54) | 0.303 | 3/29(10.34) | 2/32(6.25) | 0.564 |
| Were the prior distributions reported? | |||||||
| Noninformative prior | 18/31(58.06) | 9/13(69.23) | 10/17(58.82) | 0.564 | 8/16(50.00) | 10/15(66.67) | 0.355 |
| Informative prior | 3/31(9.68) | 1/13(7.69) | 2/17(11.76) | 0.717 | 1/16(6.25) | 2/15(13.33) | 0.512 |
| Vague prior | 10/31(32.26) | 4/13(30.77) | 6/17(35.29) | 0.798 | 7/16(43.75) | 3/15(20.00) | 0.164 |
| Not specified | 30/61(49.18) | 13/26(50.00) | 9/26(34.62) | 0.266 | 13/29(44.83) | 17/32(53.13) | 0.521 |
| Sensitivity analysis based on priors | 4/61(6.56) | 2/26(7.69) | 2/26(7.69) | 1.000 | 2/29(6.90) | 2/32(6.25) | 0.920 |
| Was the convergence assessed? | |||||||
| Not reported | 37/61(60.66) | 15/26(57.69) | 14/26(53.85) | 0.782 | 17/29(58.62) | 20/32(62.50) | 0.759 |
| Gelman Rubin statistic | 21/24(87.50) | 11/11(100.00) | 10/12(83.33) | 0.166 | 9/12(75.00) | 12/12(100.00) | 0.070 |
| Visual plot inspection | 13/24(54.17) | 5/11(45.45) | 7/12(58.33) | 0.546 | 7/12(58.33) | 6/12(50.00) | 0.688 |
| Observation of chain mix | 2/24(8.33) | 2/11(18.18) | 0/12(0.00) | 0.131 | 1/12(8.33) | 1/12(8.33) | 1.000 |
| Was the heterogeneity in NMA assessed? | |||||||
| Precision (Tau2) | 4/61(6.56) | 0/26(0.00) | 4/26(15.38) | 0.039 | 4/29(13.79) | 0/32(0.00) | 0.031 |
| Not reported | 57/61(93.44) | 26/26(100.00) | 22/26(84.62) | 0.039 | 25/29(86.21) | 32/32(100.00) | 0.031 |
| Was the similarity in NMA assessed? | |||||||
| Not reported | 53/61(86.89) | 22/26(84.62) | 22/26(84.62) | 1.000 | 23/29(79.31) | 30/32(93.75) | 0.098 |
| Comparing patients’ or trials’ characteristics | 5/8(62.50) | 2/4(50.00) | 3/4(75.00) | 0.495 | 4/6(66.67) | 1/2(50.00) | 0.693 |
| Investigating potential effect modifying covariates | 3/8(37.50) | 2/4(50.00) | 1/4(25.00) | 0.495 | 2/6(33.33) | 1/2(50.00) | 0.693 |
| Were the inconsistency assessed? | |||||||
| Not reported | 44/61(72.13) | 20/26(76.92) | 16/26(61.54) | 0.234 | 20/29(68.97) | 24/32(75.00) | 0.603 |
| Inconsistency/incoherence factors | 4/17(23.53) | 1/6(16.67) | 2/10(20.00) | 0.873 | 3/9(33.33) | 1/8(12.50) | 0.327 |
| Hypothesis test | 8/17(47.06) | 2/6(33.33) | 6/10(60.00) | 0.317 | 4/9(44.44) | 4/8(50.00) | 0.824 |
| Node-splitting analysis | 5/17(29.41) | 3/6(50.00) | 2/10(20.00) | 0.225 | 2/9(22.22) | 3/8(37.50) | 0.503 |
| Was the publication or reporting bias assessed? | |||||||
| Comparison-adjusted funnel plot | 1/61(1.64) | 0/26(0.00) | 1/26(3.85) | 0.317 | 0/29(0.00) | 1/32(3.13) | 0.341 |
| Not reported | 60/61(98.36) | 26/26(100.00) | 25/26(96.15) | 0.317 | 29/29(100.00) | 31/32(96.87) | 0.341 |
| Additional analyses: | |||||||
| Was a sensitivity analysis performed? (Yes) | 24/61(39.34) | 10/26(38.46) | 12/26(46.15) | 0.578 | 12/29(41.38) | 12/32(37.50) | 0.759 |
| Was subgroup analysis or meta-regression performed? (Yes) | 12/61(19.67) | 3/26(11.54) | 7/26(26.92) | 0.163 | 7/29(24.14) | 5/32(15.63) | 0.407 |
| Was the GRADE used? (Yes) | 3/61(4.92) | 1/26(3.85) | 2/26(7.69) | 0.556 | 2/29(6.90) | 1/32(3.13) | 0.500 |
*9 studies published in journals with no associated impact factor.