Skip to main content

Advertisement

Table 3 Bivariate qualitative analyses of all investigated parameters. Significant correlations (P < 0.05) are highlighted by bold digits

From: The fading of reported effectiveness. A meta-analysis of randomised controlled trials

  Reported effect size* Baseline Study size Treatment group
  Ph R P-value Ph R P-value Ph R P-value Ph R P-value
Year of Publication P: -0.5360 <.0001 P: -0.6056 <.0001 P: 0.2444 0.0517 P: -0.4092 0.0008
  A: 0.0305 0.8618 A: -0.1518 0.3840 A: 0.1237 0.4789 A: -0.3207 0.0603
  T: -0.5043 <.0001 T: -0.3955 0.0004 T: 0.4512 <.0001 T: -0.4513 <.0001
  L: -0.4642 0.0074 L: -0.3413 0.0559 L: -0.0346 0.8508 L: -0.3167 0.0774
Reported effect size 1.00 P: 0.2591 0.0387 P: -0.1899 0.1327 P: 0.1598 0.2073
     A: -0.1454 0.4045 A: 0.0381 0.8281 A: -0.4339 0.0092
     T: 0.6949 <.0001 T: -0.1364 0.2432 T: 0.1565 0.1801
     L: 0.8745 <.0001 L: 0.2206 0.2251 L: 0.0289 0.8753
Baseline     1.00 P: -0.2897 0.0202 P: 0.0872 0.4933
        A: -0.0868 0.6201 A: 0.0848 0.6282
        T: -0.1923 0.0983 T: 0.0686 0.5587
        L: 0.2673 0.1392 L: -0.1069 0.5604
Study size        1.00 P: 0.0566 0.6568
           A: 0.1106 0.5269
           T: -0.2723 0.0181
           L: -0.3557 0.0457
  1. Abbreviations: Ph, pharmaceutical; T, Timolol; L, Latanoprost; P, Pravastatin; A, Atorvastatin.
  2. * Reported effect size for Pravastatin and Atorvastatin is measured in relative terms (%), for Timolol and Latanoprost in absolute terms (mmHg).
  3. As the parameter treatment group only has the possibilities control group (= 0) or experimental group (= 1), point biserial correlation was used.
  4. R: Pearson correlation coefficient.