Tsuji 2009.
| Methods | Study design: C‐RCT Unit of allocation: physician Stratification by: healthcare unit size and geographic location Type of comparison: PEM only vs. nothing
|
|
| Participants | Physicians Clinical speciality: general practice/family medicine Level of training: fully trained Setting/country: general practice/Brazil |
|
| Interventions | The PEM studied in this report was a depression‐specific guide, adapted from rigorous previously published guidelines, which provided brief and objective educational information regarding the effects of depression on patient daily living, strategies for improving adherence to treatment, and guidelines for the therapeutic management planning using standardised antidepressants in primary care | |
| Outcomes | 1 process outcome: preccscription of an antidepressant at the first appointment with the clinician 1 patient outcome: clinical remission (proportion of patients with depression severity of less than 8 points on Hamilton Rating Scale for Depression Severity) |
|
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | No information was provided |
| Allocation concealment (selection bias) | Low risk | COMMENT: the unit of allocation was by physician and allocation was performed on all units at the start of the study |
| Baseline characteristics similar (selection bias) | Low risk | Quote, pg. 223: "clinician and patient baseline characteristics were comparable in the experimental and control groups (Tables 1 and 2)" |
| Baseline outcome measurements similar (selection bias) | Unclear risk | Baseline outcomes were not reported for this RCT |
| Incomplete outcome data (attrition bias) All outcomes | Low risk | Quote, pg. 223: "dichotomous end points (withdrawals, appropriate treatment and 16‐week clinical remission) were analyzed using the adjusted chi‐square approach." Withdrawals were quantified by group and reason, quote. pg. 223: "There were a total of 36 study withdrawals, 13 (10.8%) in the intervention arm and 23 (20.2%) in the usual care arm (intracluster coefficient correlation = 0.032, P = 0.153). Nine subjects (7.5%) in the intervention arm and 19 (16.7%) in the usual care arm withdrew (P = 0.122). Eight subjects, four (3.3%) in the intervention arm and four (3.5%) in the usual care arm, worsened and were withdrawn (P = .949)" |
| Blinding of outcome assessment (detection bias) All outcomes | Low risk | Quote, pg. 222: "investigators were blind to the treatment assignment of the clinicians and to which clinician the patient was assigned" and, "at 16‐week depression severity, as measured by the HAM‐D scale, was evaluated at a mental health facility by two independent evaluators who were blind to treatment allocation" |
| Contamination protection (contamination bias) | Low risk | Quote, pg. 222: "to avoid cross‐contamination of clinicians, sensitization of patients and for administrative reasons eight clinicians were stratified by basic healthcare unit size and geographical area and randomized to use either usual care or a treatment guide in treating depression" |
| Selective reporting (reporting bias) | Low risk | All relevant outcomes in the methods section were reported in the results section |
| Other bias | Low risk | There was no evidence of other risks of bias |