Table 3. Linear regression derived estimates and 95% CI with mean overall quality score on a modified percentage scale as the dependent variable for psychiatry RCT abstracts.
| Characteristics | Univariate analysis, estimate 95% CI |
Multivariate analysis, estimate 95% CI |
|---|---|---|
| Year of publication | ||
| Pre-CONSORT | Reference | Reference |
| Post-CONSORT | 6.96 (4.66, 9.27)*** | 7.3 (5.2, 9.4)*** |
| Intervention | ||
| Pharmacological | Reference | Reference |
| Psychological | -3·4 (-8·34, 1·55) | -4.74 (-7.75, -1.72)** |
| Pharmacological and psychological | -0·55 (-8·43, 7·33) | -3.16 (-6.47, 0.14) |
| Others | -2·13 (-7·08, 2·81) | -0.8 (-4.71, 3.1) |
| Journal type | ||
| High-impact general medical journals a | Reference | Reference |
| Psychiatry journals b | -25·3 (-32·18, -18·42)*** | -19.94 (-24.59, -15.28)*** |
| Impact factor | ||
| Less than 5 | Reference | Reference |
| 5–10 | 10·78 (3·58, 17·99)** | 3.31 (-2.61, 9.23) |
| More than 10 | 11·98 (4·37, 19·59)** | -2.32 (-9, 4.35) |
| Number of authors | ||
| Less than 4 | Reference | |
| 4–7 | 3·56 (-4·32, 11·45) | 2.96 (-0.63, 6.55) |
| More than 7 | 6·85 (-1, 14·7) | 5.21 (1.52, 8.9)** |
| Funding source | ||
| No funding or none reported | Reference | Reference |
| Government/peer reviewed/cooperative groups | 7·39 (-2·98, 17·76) | -0.18 (-3.87, 3.5) |
| Industry | 14·83 (3·52, 26·15)* | 2.21 (-1.87, 6.29) |
| Both | 17·94 (5·64, 30·23)** | 3.12 (-2.18, 8.41) |
| Region of publication | ||
| Europe | Reference | Reference |
| North America | -0·68 (-8·01, 6·66) | -0.05 (-2.46, 2.37) |
| Asia | -0·11 (-4·3, 4·07) | 1.75 (-2.12, 5.62) |
| Others | -4·63 (-14·03, 4·77) | 0.29 (-4.11, 4.68) |
| Centers | ||
| Single center | Reference | Reference |
| Multicenter | 10·15 (6·59, 13·7)*** | 3.7 (1.24, 6.16)** |
| Trial outcome | ||
| Unclear | Reference | Reference |
| Positive | 12·65 (6·15, 19·15)*** | 7.86 (4.22, 11.5)*** |
| Negative | 8·87 (2·12, 15·63)* | 5.96 (2.2, 9.71)** |
| Abstract structure | ||
| Unstructured | Reference | Reference |
| Structured | 10·59 (5·31, 15·88)*** | 9.08 (4.17, 13.99)*** |
| Sample size | ||
| ≤100 | Reference | Reference |
| >100 | 6·29 (2·57, 10)** | 1.96 (-0.45, 4.38) |
| CONSORT endorsement | ||
| No | Reference | Reference |
| Yes | 5·8 (0·66, 10·94)* | -0.19 (-3.58, 3.21) |
| Word count limit | ||
| < 250 | Reference | Reference |
| ≥250 or no word limit | 0·51 (-3·57, 4·59) | 3.14 (0.22, 6.05)* |
a BMJ, JAMA, N Engl J Med, and Lancet.
b Acta Psychiatr Scand, Addiction, Am J Psychiatry, Arch Gen Psychiatry (renamed JAMA psychiatry in 2013), Biol Psychiatry, Bipolar Disord, Br J Psychiatry, Eur Neuropsychopharmacol, J Child Psychol Psychiatry, J Clin Psychiatry, J Neurol Neurosurg Psychiatry, J Psychiatry Neurosci, Int J Neuropsychopharmacol, Neuropsychopharmacology, Psychol Med, Psychoneuroendocrinology, Psychother Psychosom, Schizophr Bull.
*P<0.05
**P<0.01
***P<0.001.