Laboratory Testing in Primary Care Frequently Triggered Follow-up Care
A large cluster randomised trial in Belgian primary care found that laboratory testing often set off a chain of downstream activity, even when clinicians ordered tests for routine check-up purposes. Across 10 270 laboratory panels in 9683 patients, nearly one in three tests led to some form of additional diagnostic or therapeutic action. In practical terms, an order placed to clarify a symptom, monitor a chronic condition, or simply review a patient’s status frequently became the starting point for more testing, more referrals, or changes in treatment.
The study examined what happened after the first laboratory exam rather than the test itself. That framing matters because the clinical burden of testing is not limited to the blood draw or the result on a screen. Abnormal or borderline findings can trigger imaging, specialist referral, endoscopy, repeat blood work, or medication changes, all of which affect patient experience, resource use, and the likelihood of further incidental findings. In primary care, where laboratory ordering is common and often broad, those downstream consequences can accumulate quickly.
Cluster Randomised Trial Across Belgian General Practices
Investigators embedded the trial in 72 primary care practices involving 272 general practitioners in Belgium. Practices were randomly assigned to either a laboratory ordering system with an integrated clinical decision support system or to usual ordering software. The intervention ran for 6 months, followed by 6 months of follow-up.
The cohort reflected a typical primary care population. Women made up 55.1% of participants, and the mean age was 56.5 years. The trial included adults who consulted participating general practitioners during the study period and required a laboratory examination. In that setting, check-up panels were especially relevant because they are often ordered to screen broadly rather than to answer one tightly defined clinical question.
A single laboratory order can launch a cascade when the result lands outside the expected range. Even modest departures from normal often prompt the next clinical step, particularly in patients with multiple comorbidities or in those whose baseline risk profile is already uncertain.
Nearly One in Three Panels Led to Additional Action
The main exploratory outcome was downstream activity after the initial laboratory exam. Across all panels, 19.7% led to further diagnostic procedures, with a 95% confidence interval of 18.9% to 20.5%. 19% led to treatment changes, with a 95% confidence interval of 18.2% to 19.7%. Those figures capture separate downstream pathways, and together they show that laboratory testing did not simply end with interpretation of the result. It frequently altered the clinical plan.
Check-up laboratory exams behaved in a similar way. They produced diagnostic downstream activity in 17.5% of cases, with a 95% confidence interval of 13.8% to 21.2%, and treatment changes in 18.9%, with a 95% confidence interval of 13.9% to 23.9%. The similarity between general laboratory testing and check-up panels is clinically important. Screening-style orders are often assumed to be lower yield or less disruptive, yet they still led to substantial downstream work.
Decision Support Reduced Several Downstream Pathways
Use of the clinical decision support system was associated with fewer downstream referrals and fewer treatment changes. Referrals fell by 2.4% (95% CI -4.2% to -0.6%; p=0.0008). Imaging and endoscopy also declined by 0.9% (95% CI -1.6% to -0.1%; p=0.0026). Treatment changes decreased by 5.4% (95% CI -9.5% to -1.2%; p=0.01).
Those effects point to a plausible mechanism: decision support may have helped general practitioners choose more targeted laboratory panels, avoid low-value follow-up, or interpret results in a way that limited reflex escalation. In primary care, where the pretest probability of serious disease is often lower than in specialty settings, this kind of ordering guidance can matter. Less indiscriminate testing can mean fewer borderline abnormalities, fewer cascades of uncertainty, and fewer interventions driven by noise rather than disease.
The reduction in treatment changes deserves particular attention. Medication adjustments are common after laboratory testing, but not every abnormal result carries equal clinical weight. A decision support tool that aligns testing with evidence-based order sets may reduce changes prompted by findings that would not have altered outcomes, while still allowing clearly actionable abnormalities to lead to treatment.
Check-up Panels Were Not Harmless Background Noise
Check-up laboratory testing is sometimes treated as a low-stakes exercise in reassurance. The data tell a different story. Even when panels were ordered for follow-up or screening-like purposes, downstream procedures and treatment changes remained frequent. That pattern suggests that the downstream burden is built into the structure of laboratory medicine itself: once a result exists, clinicians feel compelled to respond, and patients often enter a new diagnostic pathway whether or not the original question was narrow.
For general practitioners, this has immediate implications. Lab orders are not isolated events; they are clinical decisions that can widen the diagnostic funnel. The more broadly a panel is ordered, the greater the chance that an incidental or borderline value will generate more tests, more appointments, and more therapeutic modification. In a busy primary care environment, that can reshape workload long after the original consultation has ended.
What This Trial Adds to Primary Care Stewardship
This was the largest study to date examining downstream activities after laboratory testing, and it extends the focus beyond test utilization alone. Many efforts to improve laboratory stewardship measure how often tests are ordered, but this trial shows why downstream consequences deserve equal attention. A test that seems efficient at the point of ordering may still create substantial follow-up work after the result returns.
The cluster design strengthens the real-world relevance of the findings because the intervention was delivered at practice level rather than in a highly controlled laboratory environment. The data also show that decision support can influence not just ordering behavior but the cascade that follows. Even modest absolute reductions in referrals, imaging, endoscopy, and treatment changes may matter when multiplied across large primary care populations.
Still, the outcome here was exploratory, and the publication does not claim that every downstream action was unnecessary. Some referrals, procedures, and treatment changes were undoubtedly appropriate and clinically important. The harder issue is separating beneficial follow-up from avoidable cascade, and this trial shows that the boundary is wide enough to matter in everyday practice.
The registered trial identifier is NCT02950142. The unresolved question is not whether laboratory testing triggers downstream activity; it clearly does. The more useful question for future work is which ordering strategies reduce low-value cascades while preserving the diagnostic yield that primary care still needs.