The article being critiqued is “The Intersection of Violence, Substance Use,
Depression, and STDs: Testing of a Syndemic Pattern Among Patients Attending an Urban STD Clinic” by Senn, Carey, and Vanable (2010). This critique will address data collection and measurement; procedures; data analysis; findings; and summary assessment. The primary objective of the study was to examine the significant rates of health problems and psychosocial issues which appear to occur in patients who present at sexually transmitted disease (STD) clinics; the authors report that these patients are predominantly – the authors use the word “disproportionately” – “urban, low-income, and racial/ethnic minorities” (Senn, Carey, & Vanable, 2010, p. 614). Using these rates the authors intend to identify whether or not the problems are comorbid and whether or not they suggest a syndemic.
The data collection and measurement section reveal a significant number of variables (which are also detailed in the article’s title). These variables are the predictors being assessed to determine comorbidity and the presence of a syndemic. The variables include childhood sexual abuse (CSA), depressive symptoms, substance use, intimate partner violence (IPV), and syndemic. Each of these variables was tested in its own way using either established surveys or measures, such as the Center for Epidemiologic Studies Depression Scale (Short Form) (CESD) to determine depression symptoms. The authors more than adequately describe the measures they use and why those measures were chosen, providing information which allows the reader to determine that the measures were good choices for the study’s purpose and population.
The study provides some evidence regarding the reliability and validity of the measures. For example, the reliability and validity of the CESD have been established by previous research, which the authors indicate in the article. However, the measure for substance use focusing on marijuana is predicated on the idea that “research has found that any marijuana use, regardless of the level of use, was associated with sexual risk behavior” (Senn, Carey, & Vanable, 2010, p. 616). While this may in fact be accurate, it does not necessarily account for legitimate medicinal marijuana use, therefore calling into question the reliability and validity of this particular measure.
In terms of procedure(s), this study did not focus on an intervention. Instead, it sought to gather and collate data to determine comorbidity and syndemic. Based on the procedures described in the article, it appears that bias was minimized. There were clearly defined criteria for participation in the study. Data was collecting using a computerized survey; this method may contribute to bias in so much as the authors collected information regarding education level. Discomfort regarding computer use might have influenced participants’ responses or willingness to participate, regardless of their ‘objective’ eligibility to participate. Another concern is the inherent dangers of using self-reported behavior. While the researchers had access to the participants’ medical records and could therefore confirm or validate certain aspects of the participants’ information, this does not necessarily correct for potential problems which can emerge from this kind of data. In terms of the staff responsible for collecting data, the authors indicate that a “trained research assistant (RA)” screened potential participants (Senn, Carey, & Vanable, 2010, p. 615). This suggests that the RA had received training, but it is not clear what training that was or whether it was adequate.
Based on the data analysis section, it appears that appropriate statistical methods were used. These methods include odds ratios (adjusted and unadjusted); logistic regressions; and predictors were reported grand mean centered (Senn, Carey, & Vanable, 2010). It is not exactly clear whether the most powerful analytic method used, but it is clear that the authors conducted several different logistic regressions with regard to different pairwise interactions. This approach has the potential to control for confounding variables. Based solely on the data analysis section, it is not clear whether the hypotheses were supported; this section only addressed the various methods used. The results/findings section addressed whether the hypotheses were supported.
In terms of the findings, the authors do present information about statistical significance. They present extensive details on the percentages found for each variable relative to other variables and what they mean in terms of additive and interactive effects on the larger context of syndemic. However, no information regarding effect size and the precision of estimates (including probability statistics) was presented. Ultimately, the findings of the study support the idea that syndemic is present. According to the authors, their study is the first of its kind to address an STD/HIV syndemic, which they did identify. Subsequent studies of similar syndemics, with similar populations, have since echoed the findings of this study. For example, Illangasekare, Burke, Chander, and Gielen (2013) found similar syndemic results, though their study focused exclusively on women; this study population came from low-income urban areas.
Despite this study’s limitations, some of which can be attributed to it being the first of its kind, the study’s findings appear valid. This reader has confidence in the truth value of the results, though this reader also has some concerns regarding the self-reported nature of the data and the ways substance use was defined/valued. But as a piece of evidence in the context of the studied issues, the article does offer useful information in general. It also contributes meaningful evidence that can be used in nursing practice, as well as information that could be useful to the nursing discipline. This evidence focuses on the implications of certain factors (e.g., CSA, depressive symptoms, and substance use) and how they relate or contribute to syndemic. This study has implications for understanding health risk factors and how they interact.
