A review of ssri-induced indifference
Download 203.85 Kb.
|
| Cross-sectional Studies
There were twelve (12) cross-sectional studies which have formed the basis for the consensus in the various reviews for the prevalence of SSRI-induced emotional blunting A key problem for many of them is the selection bias inherent in the recruitment process. For example, in Goodwin et al.12 seven thousand nine hundred and sixty-six (7,966) invitations sent with one thousand and four (1004) total subjects agreeing to the study, conducted under a market research protocol. Another example is Price et al.19 who conducted a cross-sectional study with nineteen (19) out of a sample of thirty-eight (38) patients recruited from a recruitment poster. Opbroek et al.20 conducted a cross-sectional study involving fifteen (15) patients reporting sexual dysfunction where 80% also described significant blunting on the Laukes Emotional Intensity Scale Questionnaire, but the selection bias compromises any conclusions. Hughes et al.1 conducted a qualitative survey of internet forum postings, where users are not blinded from each other’s posting, with a serious risk of selection bias and the spreading of ideas. The self-reporting bias is evident in the very high percentages reported in the online survey of antidepressant users for adverse effects, conducted by Read et al.15 for disparate outcomes such as “Suicidality” (50%), “Addiction” (40%) and “feeling foggy or detached” (60%). A problem inherent in all cross-sectional studies is the impossibility of knowing the temporal relation of the outcome to the exposure (i.e. SSRI treatment). A study that promised to study the temporal relation more accurately was the cohort study by Dell’Osso et al.24 of a hundred and thirty-five (135) patients attending an Obsessive Compulsive Disorder (“OCD”) clinic. They were followed up at monthly intervals for a mean period of three (3) years. Unfortunately, no analysis was made of the changes over time except a cross-sectional finding that 14% of patients had scores positive for apathy. Some studies attempted to control for confounding factors, such as in Yee et at.17 where depressive symptoms and educational levels were found to be associated with differences in anhedonia scores and the authors attempted to adjust for these. Still, being cross-sectional and without a baseline prior to commencement of treatment, it is impossible to conclude causality and the authors admit as much. In another case control by Marazziti et al.18 comparing one hundred and ninety-two (192) patients who were either on tricyclic antidepressants or SSRIs for impairments in “loving feelings” measured on a Sex Attachment and Love Questionnaire devised by the authors, the problem of multiple comparisons is poorly handled with the authors picking the items amongst the forty-item self-report that showed a positive result, without any correction of this. The problem of measuring subjective outcomes are poorly dealt with in most of these studies, with few studies describing the methodology of applying the instruments. One study that did discuss the limitations of the measurement of apathy and their validity was the case control study by Wongpakaran et al.13 A Geriatric Hospital Database of elderly depressed patients who had received antidepressants was divided into two (2) groups depending on antidepressant use at discharge, an SSRI group and a non-SSRI group. Apathy scales developed by the authors were selected from the Geriatric Depression Scale and the Hamilton Rating Scale. SSRI use was not a predictor of apathy at admission while it was at discharge with an odds ratio of 1.9 (1.14 – 3.17) compared to the non-SSRI group p=0.029. The authors concluded that apathy was greater in patients on SSRIs. All apathy scores at discharge were less than that at admission. Other confounding factors associated with apathy were analyzed, such as comorbid medical problems and medications, age and length of stay with length of stay one of the predictor variables for apathy (p=0.011) Download 203.85 Kb. Share with your friends:
|