On 2013 Oct 18, James C Coyne commented:

The authors had a responsibility to inform readers that patients assigned to the control group had to pay for any care related to diagnosis and treatment of depression. In contrast, patients assigned to the intervention group received free care. More care resulting from care being free might readily explain extraordinary differences between the two groups in receipt of care and in the primary outcome.

For an extended discussion, see my PLOS blog post at Mind the Brain. Coordinating depression treatment from afar: Are results credible?

http://blogs.plos.org/mindthebrain/2013/03/12/coordinating-depression-treatment-from-afar-are-results-credible/

The letter posted below was rejected from JAMA Internal Medicine, eliminating the possibility for post publication correction of the limitations of peer review of this paper.

Effect size in depression trial could be due to inadequacies of control treatment: Comment on Davidson and colleagues (2013) Davidson and colleagues [1] claimed benefits for a collaborative care (CC) intervention for depression that exceed not only previous CC trials, but also effect sizes for a variety of interventions for depression. Strong claims often later prove to be exaggerated or simply false [2] and deserve special scrutiny. We should keep in mind that effect sizes observed in trials are not attributes of interventions, but of comparisons between interventions and control groups. Large effect sizes can simply represent the exceptionally poor outcomes of control groups. Davison et al.’s underspecified “routine” care could simply have been inadequate care. The authors failed to acknowledge that patients in the control group had to pay for any depression treatment, whereas it was provided free to the patients in the intervention group. This might explain that the number of patients in the “routine” care group who received a new prescription of antidepressants increased by only two, versus ten in the intervention group. Similarly, the number of patients in the routine care group that received psychotherapy increased only by seven, versus an increase of 42 in the intervention group. The low rate of increased treatment occurred in the control group despite providers having been informed of patients’ depression scores. Patients were designated as “depressed” based on a self-report questionnaire. Thus, we cannot determine the extent to which patients with heightened depressive symptoms but failing to meet formal criteria for major depression were appropriately not having treatment initiated or inappropriately treated. Overall, we cannot determine whether active treatment or the mere attention and support and awareness of treatment being available free were associated with the greater improvement in the intervention group. Moreover, most patients identified as “depressed” in the intervention group were not in remission at follow up. Difficulty interpreting results could have been anticipated at the time of the study’s design. In short, results of this trial are insufficient to encourage a more ambitious trial with the same basic design, because of a lack of demonstration that any particular elements of centralized care management account for the group differences in improvement in depression that were observed, rather than inadequacies in the care provided to the control group.

1. Davidson, KW, Bigger, JT, Burg, MM, Carney, RM, Chaplin, WF, Czajkowski, S, Dornelas, E, Duer-Hefele, J, Frasure-Smith, N, Freedland, KE; Haas, DC; Allan S. Jaffe, AS,, Ladapo, JA,; Lespe´rance, F, Medina, V, Newman, JD, Osorio, GA Parsons, F, Schwartz, JE, Shaffer, JA Shapiro, PA,. Sheps, DS, Vaccarino, V, Whang, W, Ye, S. Centralized, Stepped, Patient Preference–Based Treatment for Patients With Post–Acute Coronary Syndrome Depression CODIACS Vanguard Randomized Controlled TrialCODIACS Vanguard RCT. JAMA Internal Medicine, 2013, 1-8.

2. Ioannidis, J. P. (2005). Why most published research findings are false. PLoS Medicine,

On 2013 Oct 18, James C Coyne commented:

The authors had a responsibility to inform readers that patients assigned to the control group had to pay for any care related to diagnosis and treatment of depression. In contrast, patients assigned to the intervention group received free care. More care resulting from care being free might readily explain extraordinary differences between the two groups in receipt of care and in the primary outcome.

For an extended discussion, see my PLOS blog post at Mind the Brain. Coordinating depression treatment from afar: Are results credible?

http://blogs.plos.org/mindthebrain/2013/03/12/coordinating-depression-treatment-from-afar-are-results-credible/

The letter posted below was rejected from JAMA Internal Medicine, eliminating the possibility for post publication correction of the limitations of peer review of this paper.

Effect size in depression trial could be due to inadequacies of control treatment: Comment on Davidson and colleagues (2013) Davidson and colleagues [1] claimed benefits for a collaborative care (CC) intervention for depression that exceed not only previous CC trials, but also effect sizes for a variety of interventions for depression. Strong claims often later prove to be exaggerated or simply false [2] and deserve special scrutiny. We should keep in mind that effect sizes observed in trials are not attributes of interventions, but of comparisons between interventions and control groups. Large effect sizes can simply represent the exceptionally poor outcomes of control groups. Davison et al.’s underspecified “routine” care could simply have been inadequate care. The authors failed to acknowledge that patients in the control group had to pay for any depression treatment, whereas it was provided free to the patients in the intervention group. This might explain that the number of patients in the “routine” care group who received a new prescription of antidepressants increased by only two, versus ten in the intervention group. Similarly, the number of patients in the routine care group that received psychotherapy increased only by seven, versus an increase of 42 in the intervention group. The low rate of increased treatment occurred in the control group despite providers having been informed of patients’ depression scores. Patients were designated as “depressed” based on a self-report questionnaire. Thus, we cannot determine the extent to which patients with heightened depressive symptoms but failing to meet formal criteria for major depression were appropriately not having treatment initiated or inappropriately treated. Overall, we cannot determine whether active treatment or the mere attention and support and awareness of treatment being available free were associated with the greater improvement in the intervention group. Moreover, most patients identified as “depressed” in the intervention group were not in remission at follow up. Difficulty interpreting results could have been anticipated at the time of the study’s design. In short, results of this trial are insufficient to encourage a more ambitious trial with the same basic design, because of a lack of demonstration that any particular elements of centralized care management account for the group differences in improvement in depression that were observed, rather than inadequacies in the care provided to the control group.

1. Davidson, KW, Bigger, JT, Burg, MM, Carney, RM, Chaplin, WF, Czajkowski, S, Dornelas, E, Duer-Hefele, J, Frasure-Smith, N, Freedland, KE; Haas, DC; Allan S. Jaffe, AS,, Ladapo, JA,; Lespe´rance, F, Medina, V, Newman, JD, Osorio, GA Parsons, F, Schwartz, JE, Shaffer, JA Shapiro, PA,. Sheps, DS, Vaccarino, V, Whang, W, Ye, S. Centralized, Stepped, Patient Preference–Based Treatment for Patients With Post–Acute Coronary Syndrome Depression CODIACS Vanguard Randomized Controlled TrialCODIACS Vanguard RCT. JAMA Internal Medicine, 2013, 1-8.

2. Ioannidis, J. P. (2005). Why most published research findings are false. PLoS Medicine,