The sample consisted of 264 employers who responded at 12 month follow-up (81.2% of the baseline sample). The organizational characteristics of responding employers are described in Table 1.

Although the EB and UC cohorts contained companies of statistically comparable size (p > 0.10), large employers were less likely to complete 12 month follow-up (p < 0.05). Thus, we controlled for company size in the comparisons we present in this manuscript. At 12 months, 29.2% (n = 38) of EB employers reported internal discussion of depression products by email exchange, telephone call, in-person meeting or group meeting with other employees of the organization only, compared to 33.1% (n = 43) of UC employers (p = 0.82).

Employers reporting internal discussion were asked about arguments made in support of and against depression product purchase. As Table 2 demonstrates, EB and UC employers reported their discussions included generally similar arguments in support of purchasing a depression product. Given that the EB presentation emphasized the scientific evidence demonstrating that depression products increase productivity and provide a return on investment, it was not surprising that these arguments occurred more often in EB than UC discussions.

One EB employer indicated that “associates who are directly or indirectly impacted by depression need some type of assistance to help them through the depression” and then went on to say that “in addition to direct health care costs…there can be lower productivity for the associate and co-workers.” Another EB employer indicated that offering a depression product “benefits us as the employer because the employee can get the help they need to get well and ultimately return to work as a major contributor.” A third EB employer succinctly stated that having a product would “keep overall costs down and production up.” UC employers echoed these arguments when they noted a “healthier workforce and less absenteeism (equals) greater productivity,” and the importance of “(keeping) employees healthy” in order to “keep medical costs from going up due to untreated depression.”

As Table 3 demonstrates, the arguments made against a depression product purchase were generally comparable in EB and UC discussions.

One EB employer reflected the information in the presentation by noting “this [purchasing a depression product] is a no brainer.” Contradicting information in the presentation, a second EB employer commented that “depression products are so person to person specific [that] it is impossible to gauge what will be helpful and what will not” and a third commented “does anything work?” Many UC employers, who had not been educated about the product’s return on investment indicated that “cost” was an argument made against depression product adoption without elaborating; however, one employer considered the benefits of a depression product to be a “poor performer [with regard to] costs.”

All employers not currently in the purchasing process were asked what would trigger their organization’s purchase of a depression product. As Table 4 demonstrates, EB and UC employers reported similar triggers might stimulate the purchase of a depression product. The most commonly reported trigger in both groups was a visible episode of depression in the company, often referred to as a “poster child.” Other common triggers were company-specific data on lost productivity, (further) increases in healthcare costs, management change and economic/financial improvement.

One EB employer noted: “huge increase in costs related to depression—or a tragedy like the suicide of one of our employees due to depression.” Another responded: “obvious loss of productivity clearly tied to incidents of prolonged depression.” Despite having received an estimate of the return on investment for his/her company for a depression product, a third EB employer said: “the demand for the product and impact on productivity must have been identified and the cost of now providing the product estimated to be higher than the cost.” One UC employer responded: “a trend where employees are absent more frequently using short-term disability pay with the diagnosis of depression.” Another said: “We are 100% self-insured. Until we recover financially, we will not be able to purchase any product. It is purely financial.” A third said “depression would have to be in our top five chronic conditions for us to purchase a depression depression management product.”

All employers were asked about how return on investment impacted benefit decision-making. Reflecting the organizational comparability of EB and UC companies, it is not surprising the two groups reported that return on investment had a similar impact on decision-making about additional health benefits. As shown in Table 5, close to half of employers in each group reported return on investment had a large impact, while 20% indicated it had no impact or failed to answer the question.

Approximately 30% of employers in both the EB and UC groups raised the idea that their companies should consider depression products in internal discussions in their organizations. From the glass half empty perspective, two out of every three EB employers failed to ‘bring home’ any ideas from a two hour presentation. While the presentation contained one role play between the presenter and the audience, it is possible that future interventions could be strengthened by utilizing more targeted role plays to increase employer skill/comfort to lead these discussions. However, we suspect that EB employers who chose to keep silent speak volumes about their company’s interest in improving depression treatment in the workforce. Earlier interest may have been “knocked off the radar screen” by the considerable retooling that healthcare reform required from already overstretched health benefit professionals. From the glass half full perspective, the intervention appears to have equipped committed health benefit professionals to be stronger advocates for depression products. Almost half of EB discussions noted that depression products can improve productivity; however, that message may be already circulating among health benefit professionals, as one quarter of UC discussions recognized that as an advantage. Also encouraging is that almost half of participating employers recognized that return on investment had a large influence on new benefits. Our anecdotal data indicate that future research is needed to deconstruct their understanding of return on investment and the level of evidence necessary to establish it.

Employers provided unanticipated perspectives on company changes that would trigger the purchase of a depression product. While disappointing, it probably should not be surprising that a visible episode of depression in the workforce (“poster child”) was the most often reported trigger. Visible episodes of depression are rarely evident in the workplace, because depression is more likely to erode performance “in the cubicle” than in dramatic venues. More surprising was the finding that company-specific data on poor performance could act as a trigger. Rather than accept findings from “somebody else’s company,” employers may believe that their company is not beset by productivity loss when employees become depressed. While this may be true in some circumstances, national studies [8,9] suggest that reduced productivity is the rule, not the exception.

These findings are limited by the purposive sample we studied. Employers represent a group of potentially early innovators with sufficient interest in the topic to volunteer to participate in a longitudinal study. Because it should be easier to persuade this group than a nationally representative sample, we do not think this limitation reduces our certainty about the challenges of influencing employer benefit decisions using group presentations. Thus, we encourage future researchers to develop more innovative interventions to motivate employers to translate science to benefit design.