CLINICAL UPDATE · FOR HEALTHCARE PROFESSIONALS

Primary Care
Research Brief

25 September 2026

The clearest practice signal this week is a larger role for intravenous iron in selected pregnant patients. Two further studies reinforce cautious steroid prescribing and show that a text link alone does not reliably improve reproductive-health discussions.

23 SEPTEMBER · PEER-REVIEWED SYSTEMATIC REVIEW OF RANDOMISED TRIALS

Intravenous iron deserves earlier consideration in pregnancy

Question and result. For iron-deficiency anaemia in pregnancy, how does intravenous iron compare with oral iron? Across 29 randomised trials involving 11,771 participants, IV iron raised haemoglobin at delivery by a mean 0.59 g/dL and reduced transfusion during the delivery admission by 37% (RR 0.63, 95% CI 0.49–0.82). The association was stronger when pretreatment haemoglobin was 9 g/dL or lower (RR 0.58). Mild adverse effects were less frequent; no serious treatment reactions were reported.

Evidence strength and limitations. Moderate-to-high strength synthesis of randomised evidence with a clinically important transfusion outcome. However, formulations, thresholds and settings varied; haemoglobin and ferritin results were highly heterogeneous, 13 trials were lower quality, and serious adverse events were too uncommon to rule out rare harms. No clear improvement occurred in gestational age, birth weight or caesarean delivery.

GP implication. Oral iron remains a reasonable first option for many patients. Escalate earlier through the local maternity pathway when anaemia is more severe, oral treatment is not tolerated or ineffective, adherence is unlikely, or delivery is approaching. IV treatment needs an appropriate monitored service.

Verdict: Change escalation discussions now, but follow NICE, HSE or South African maternity protocols rather than offering routine IV iron to everyone.

Read the systematic review →

23 SEPTEMBER · PEER-REVIEWED SELF-CONTROLLED COHORT

Even short oral-steroid courses deserve caution in type 2 diabetes

Question and result. Are oral corticosteroid bursts of 14 days or less associated with serious harms in adults with type 2 diabetes? In 36,048 Taiwanese adults who received a burst, the 6–30 day period after initiation had higher rates of gastrointestinal bleeding (IRR 1.71), pneumonia (2.06) and heart failure (1.87) than the individual’s baseline period. Fracture risk was higher at 31–90 days (IRR 1.42).

Evidence strength and limitations. Moderate observational evidence using a self-controlled design and sensitivity analyses. It shows association, not causation: the illness prompting steroids may contribute to risk, adherence and lifestyle data were unavailable, absolute risks were not provided, and generalisability beyond Taiwan is uncertain.

GP implication. Do not treat a brief course as harmless. Confirm an evidence-based indication, use the lowest effective duration, consider alternatives for self-limiting conditions, anticipate hyperglycaemia and give clear infection, bleeding and heart-failure safety-netting.

Verdict: Change prescribing vigilance now; do not withhold steroids when their established benefit outweighs risk.

Read the corticosteroid safety study →

23 SEPTEMBER · PEER-REVIEWED CLUSTER RANDOMISED TRIAL

A one-way text link did not improve reproductive-health discussions

Question and result. Does texting patients a link to a reproductive decision-support tool before a primary-care visit increase discussion of pregnancy intentions or contraception? In 349 US female veterans, only 49.7% used the tool. Neither optimal shared-decision-making discussions (17.3% vs 14.4%) nor any discussion (70.8% vs 65.5%) increased significantly in intention-to-treat analyses.

Evidence strength and limitations. Randomised implementation evidence, but the trial stopped early, had about 70% power for the primary outcome, measured discussion rather than health outcomes, and involved a predominantly female-clinician Veterans Health Administration population.

GP implication. Digital prompts may support—not replace—a clinician asking respectfully about pregnancy intentions, contraception and preconception needs. If using a tool, build in explanation, reminders and an opportunity for dialogue.

Verdict: Practice unchanged. Do not invest in a stand-alone one-way text strategy on this evidence.

Read the cluster trial →

MOST WORTH READING · RANKED

Three primary papers

Window: 19–25 September 2026. All selected evidence is peer-reviewed; no preprints or conference-only findings are included. No verified major new NICE, Irish HSE or South African national primary-care guideline was identified for this window. Educational summary for healthcare professionals; local guidance and individual assessment take precedence.

Earlier briefings: 18 September · 11 September · 4 September.