AI-generated research brief — verify at source
Mailed SDM Intervention Increases FOBT Prescribing, But Selection Bias Clouds Results
A mailed shared decision-making intervention significantly increased the proportion of Swiss primary care physicians who had at least one patient tested or planning fecal occult blood testing compared to controls — about 52% more likely in the intervention group (84% vs. 56%; RR 1.52, 95% CI 1.13 to 2.04). However, a sensitivity analysis incorporating baseline data from a subset of physicians produced a non-significant estimate, raising the possibility that pre-existing differences between groups, rather than the intervention itself, partly explain the effect.
What Was Studied
The study examined whether a low-intensity, mailed intervention could shift primary care prescribing behavior in Switzerland away from colonoscopy-only recommendations and toward shared decision-making that includes faecal occult blood testing as an option. The primary outcome was the proportion of physicians who had at least one eligible patient tested or actively planning FOBT within four months of the intervention.
How It Was Studied
This was a cluster randomized controlled trial conducted within a Swiss primary care research network, with physicians randomized 1:1 to intervention or control. The intervention group received a package of mailed materials including patient decision aids, an evidence summary, an FOBT sample kit, and personalized feedback on their own CRC screening practices. The control group received none of these materials before data collection began. A total of 110 primary care physicians were randomized, of whom 83 (76%) went on to document screening decisions for 3,171 consecutive patients aged 50 to 75 — a mean age of 62, with equal representation by sex. Physicians recorded outcomes over four months following the intervention.
What Was Observed
- The primary outcome favored the intervention: physicians in the intervention group were substantially more likely to have at least one patient tested or planning FOBT compared to controls — roughly 52% more likely on a relative basis (84% vs. 56%; unadjusted RR 1.52, 95% CI 1.13 to 2.04).
- The sensitivity analysis weakened this conclusion: among the 62 physicians who had also participated in a prior data collection round, the relative risk remained similar in magnitude but was no longer statistically significant (RR 1.49, 95% CI 0.98 to 2.28). Crucially, even before the intervention, 72% of intervention-group physicians already met the primary outcome at baseline versus 49% of controls, suggesting pre-existing group differences.
- Within-physician change over time was not significant: when the analysis was restricted to comparing each physician’s own change in practice between the 2017 and 2018 data collections, there was no significant difference between intervention and control groups, though the intervention appeared to increase FOBT volume among physicians who were already prescribing it.
Why This Matters
In Switzerland, colonoscopy has been the default screening approach in primary care, meaning patients are rarely offered a choice that includes FOBT — a less invasive, guideline-endorsed option. This trial demonstrates that a simple, mailed package of materials can at least be associated with broader FOBT uptake across a primary care network, pointing toward the feasibility of low-resource interventions to promote patient choice in cancer screening discussions.
How to Read This Result
While the main analysis shows a statistically significant positive effect, the overall quality of evidence is medium at best, as the non-significant sensitivity analysis and the absence of within-physician change over time both suggest that selection bias — driven by pre-existing FOBT prescribing habits among intervention physicians — may account for a meaningful portion of the observed difference.
Limitations
The most important limitation is potential selection bias: physicians in the intervention group were already more likely to prescribe FOBT before the study began, as demonstrated by the baseline data from the 62-physician subgroup. The sensitivity analysis addressing this directly produced a non-significant result (RR 1.49, 95% CI 0.98 to 2.28), casting doubt on the causal interpretation of the primary finding. The within-physician longitudinal comparison also failed to show a significant change over time. Additionally, the intervention may have primarily increased FOBT use among physicians already practicing it, rather than converting colonoscopy-only prescribers — limiting the generalizability of the effect to the broader population of Swiss primary care physicians.