Tools available to screen for employment in health systems are diverse in content and implementation, but relatively few assess the complex nature of work. Results from our analysis suggest that employment items in screening tools that have been implemented in clinical healthcare delivery settings are underdeveloped and unclear in their intended purpose. In our content analysis, we first characterized the intent of the question, which we categorized into four Domains: Social Risk Factor, Social Need, Employment Exposure, and Legal Need. Half of the employment questions fell into the Social Need domain, which assessed respondents’ perceptions of employment as a concern, or which tied the question to an intervention. This Domain is similar to the Social Risk Factor domain, which assessed employment status in terms of whether the respondent was employed or what broad category of employment their work fell into (e.g., full-time vs part-time work) without assessing the respondents’ need for support. In both domains, the purpose behind the question was unclear. If a respondent indicates that employment is a concern, does this suggest that they are concerned about finances, health insurance, or social support? Or, on the other hand, is their employment causing physical or psychological stress? Questions that fell into the Employment Exposure and Legal Need domains were clearer in their intent and more easily linked to a core measurement purpose (e.g., occupational exposure, environmental exposure, work characteristics, legal benefits, the legality of work arrangements).
Next, we examined the question and response options together and characterized the extent to which the item captures a respondent’s employment context. Together with the question Domain, this helps us understand how actionable a given item is. With more than half of the items assessed having little to no context (Level 1), our results suggest ambiguity in how screening tool questions are being developed, applied, and evaluated. Items without any employment context (e.g., “Are you employed? yes/no”) may stifle the potential benefits that screening tools can provide and have implications for how a healthcare delivery system can intervene to address the need. Even items that provided somewhat more employment context (Level 2) did not capture information specific enough for a provider to adjust one’s care plan or connect patients with resources [15]. For example, there was not enough information in these items for providers to glean details that would aid them in discussing strategies to mitigate occupational risks based on an individual’s health status or help them to make appropriate referrals to social services based on an individual’s need. In contrast, items identified as highly contextualized, with a Level 3, captured nuances that would allow providers to address specific aspects of an individual’s work context.
Of the tools identified that were highly contextualized, more than half were implemented in specialized healthcare delivery settings like musculoskeletal clinics and obstetric clinics. Providers in specialized settings may better understand the nuances of how employment status, working conditions, and occupational exposures impact their patients’ health. As such, providers in these settings are likely more incentivized to address employment in a contextual way to 1) better understand how best to treat a patient and 2) better identify measures of association between employment characteristics and health outcomes, which may be less apparent in primary care settings. Findings may also suggest that items of a tool reflect the level of impact the providers feel they have in modifying or addressing employment characteristics. For example, items in the Employment Exposure domain were mostly developed with an occupational health lens, while items in the Social Risk Factor and Social Need domains were more general. Occupational health providers may feel better equipped to understand how employment characteristics influence health and may feel more comfortable intervening to address a patient’s need. More exploratory evidence that looks at who developed the items, the primary intent, and the perceived modifiable risk would provide further insight into the relationship between contextualization and domain.
In addition to providing more concrete guidance for intervening, highly contextualized items with a clear purpose may encourage patient engagement if patients better understand how answering social screening questions are linked to solutions. While many patients believe that screening for social needs is valuable, patients need to be convinced that the screening tool items are intentional and to understand how providers will use the information [22, 23]. Future research should focus on developing items with a keener eye towards what context to include to adequately assess relevant outcomes of interest. Centering the intent of the item and considering the applicable contexts would provide a foundation for researchers to assess the effectiveness of individual questions more adequately. Implementation science frameworks, like the health equity implementation framework [24], would be particularly useful to describe how variation in question intent or clinical healthcare delivery setting might influence the effectiveness of screening tools given a particular setting or population.
Our analysis further examined the healthcare setting in which the screening tool has been used. Basic, less contextualized items might suffice for a general adult population where a wide range of social risks or social needs may be present, and the impact of work-related exposures could be less pronounced. However, our results suggest that specific populations may benefit from screening tool items better tailored to their needs. For example, pregnant, pediatric, and Veteran populations frequently appeared in our analysis, with each having different considerations and needs for care. Individuals in populations such as these could be at risk for more acute health implications warranting work adjustments or may more often experience working conditions that directly link to health. Distinguishing between settings is especially important in generating generalizable evidence across clinical healthcare delivery systems.