The 10.4% Survey and the Six-Week Deadline
Bedford County (population 84,000) has three food-processing plants employing 2,300 people, most on rotating 8- and 12-hour shifts. The county health department and a regional hospital formed a wellness coalition after the state Behavioral Risk Factor Surveillance System reported county adult obesity at 33.8%. Plant human resources staff pulled occupational health screening records: 46% of screened employees had a BMI of 30 or higher, but only 64% of employees attended screening. The coalition ran an employee survey through shift supervisors and received 240 responses (10.4%), two-thirds from salaried day-shift staff. Two evening-shift focus groups (n=14) described 10-minute breaks, vending machines with no produce, and a line supervisor who discourages leaving the floor. The coalition chair, a plant vice president, wants to skip further data collection and launch a 10,000-step challenge in six weeks. The department epidemiologist has left and will not be replaced until the next fiscal year.
1. Critique the evidence base. Which findings can the coalition defend, and which cannot?
Read the worked answer
Separate each claim by what its sampling method can actually support.
- State BRFSS figure (33.8%): defensible as context, not as a county baseline. BRFSS is self-reported and telephone-based, so it understates BMI, and state-weighted estimates may not represent a county of 84,000. Use it to justify attention, not to measure change.
- Occupational screening (46% with BMI ≥30): the numerator is objective, but the denominator is 64% of employees. People who attend screening are plausibly those already concerned about health or required to attend by job duties. That is selection bias and it probably inflates the estimate. Presenting 46% as 'nearly half of all employees' overstates the finding.
- Employee survey (10.4%, two-thirds salaried day shift): this is a nonresponse-bias problem. The rotating-shift hourly workers the focus groups flag as highest risk are underrepresented. The survey can generate hypotheses and wording for messages; it cannot generate prevalence.
- Two focus groups (n=14): appropriate for explaining mechanisms such as break length, vending placement, and supervisor norms. Not appropriate for frequency or generalization, but they are the strongest data the coalition holds for choosing intervention targets.
What is defensible is the direction of the problem, at least one environmental driver, and a plausible organizational norm. What is not defensible is any precise prevalence claim or an inference that salaried staff represent all employees. The coalition also has not addressed whether reusing occupational health records requires a data use agreement or HIPAA authorization. Triangulating the four sources shifts the conclusion from 'employees are obese' to 'shift structure and the food environment constrain healthy behavior' - which points at different interventions.
2. What assessment activities would you complete in the next six to eight weeks, and what would each add?
Read the worked answer
Sequence the work so that each activity answers a question the current data cannot.
- Shift-change intercept survey (target n≈300) with a $5 incentive, administered at all three gates across day, evening, and night shifts. This addresses the nonresponse bias directly. Keep it to 10 items on break access, food purchasing, and interest in programs.
- Environmental audit of each plant: vending machine contents and pricing, break room seating and refrigeration, stairwell visibility, walking paths, water access, and the timing of breaks relative to the shift. This converts focus-group complaints into measurable, modifiable targets and gives the coalition something it can change without asking employees to change first.
- Key informant interviews (8 to 12) with plant managers, union stewards, line supervisors, and occupational health nurses. Purpose: assess organizational capacity, decision-making authority over vending contracts, and likely barriers - the PRECEDE-PROCEED 'administrative and policy' diagnosis.
- Secondary data request: absenteeism, workers' compensation, and short-term disability aggregates by plant and shift. These are the outcomes an employer vice president will actually act on, and they let you frame the case in management terms.
- Asset mapping of existing resources: the hospital's diabetes education staff, county parks, pharmacy, and any existing walking groups.
Before any of this, secure a data use agreement for the occupational health records and confirm how survey data will be stored. Document each source, its limitations, and its date in a short assessment report - a one-page data quality table prevents the 41% misuse problem later.
3. The chair wants to launch the challenge in six weeks. How do you respond?
Read the worked answer
Do not frame this as delaying against doing. Frame it as choosing what kind of launch the coalition can defend in twelve months.
The argument: a 10,000-step challenge is an individual-behavior intervention aimed at a population whose stated barriers are break timing, food access, and supervisor norms. If it launches without baseline data, the coalition will have no way to attribute any change, and the funder will ask for outcomes it cannot produce. The chair, as a plant vice president, also has a role conflict worth naming politely - a challenge is low-cost, visible, and requires no changes to plant operations, which may be exactly why it appeals to management.
Offer a workable compromise with real conditions:
- Launch a pilot challenge at one plant, one shift, as a formative test rather than a county-wide program, while the intercept survey and audit run in parallel.
- Build the pilot so it collects data: baseline and 12-week self-report plus a participation and step-log measure, and a short exit interview with non-completers.
- Commit to a written decision point at week 12: if the audit shows vending and break timing are the binding constraints, the next phase shifts to environmental change rather than another challenge.
- Ask the chair to sponsor the vending-contract conversation, which is the leverage only he has.
Get the coalition to record this decision in the minutes. Documenting the rationale protects the health educator when the pilot produces modest results, and it keeps assessment honest rather than performative.