PRESCRIBING GUIDANCE AND DISPENSING TIMING CHECK Checked 4 November 2026. Observation window: January 2024–December 2025. FINDING There was a new Scottish antidepressant guide in August 2024 and a specific East Region formulation amendment in February 2025. We did not identify a new blanket citalopram-to-sertraline policy in the dated sources reviewed. Both national share movements were already visible before August 2024. This rules out treating publication as the start of the observed movement, but does not rule out earlier influence, anticipatory implementation or guidance formalising existing practice. These data cannot establish why it happened or whether guidance accelerated it. WHAT CHANGED DURING THE WINDOW 1. 8 August 2024 — Scottish Government, Quality prescribing for antidepressants: guide for improvement 2024–2027. Scope: national Scottish guidance. It promotes person-centred reviews, indication recording, appropriate doses, non-drug options, and tapering or alternative treatment where appropriate. Its recommendations do not direct a blanket replacement of citalopram with sertraline. Publication is verified; local implementation dates and uptake are not. A new publication does not mean every recommendation was new; this check did not perform a line-by-line comparison against every predecessor document. Source: https://www.gov.scot/publications/quality-prescribing-antidepressants-guide-improvement-2024-2027/pages/4/ Detail: https://www.gov.scot/publications/quality-prescribing-antidepressants-guide-improvement-2024-2027/pages/6/ Alignment: the observed directions predate publication. Continuing trends afterwards are compatible with many explanations, including existing practice, changing starts/stops or repeat treatment. This comparison does not estimate the guide's effect. Could clinicians have acted before publication? Yes, that remains possible. The executive summary describes contributors already delivering antidepressant prescribing improvement work. The document also illustrates a citalopram-to-sertraline change for a person with inadequate treatment response; this is an individual case example, not a blanket switch policy. These features support treating publication as one point in an ongoing process. They do not demonstrate early circulation, which clinicians saw a draft, anticipatory changes, or a causal effect on our national figures. The targeted search did not locate a dated pre-publication circulation or implementation record. Publication date is therefore not a clean intervention date. Earlier movement cannot establish that guidance was irrelevant. Sources: https://www.gov.scot/publications/quality-prescribing-antidepressants-guide-improvement-2024-2027/pages/2/ https://www.gov.scot/publications/quality-prescribing-antidepressants-guide-improvement-2024-2027/pages/10/ 2. 5 February 2025 — East Region Formulary Committee, section 3.2.4, printed page 10. Scope: NHS Fife, Lothian and Borders. The committee recorded an amendment for licensed sertraline 50mg/5ml oral suspension to replace an unlicensed product in several pathways. It concerns formulation and licensing, rather than a general switch between antidepressants. The minutes say the website will be updated; the implementation date is not supplied. Source: https://formulary.nhs.scot/media/rl2ho0gx/erfc-05-02-25-minutes-final-update-30-04-2025.pdf Alignment: this is too narrow to explain the broad national pattern on its own. Sertraline tablets account for 97.3% of its net item increase. All liquids together account for the remaining 2.7%. This does not exclude a local effect on choice of liquid. RELEVANT GUIDANCE THAT PREDATES THE WINDOW 3. NHS Borders Treatment of Depression Pathway: explicit document update April 2021, review due April 2023. Page 4 already lists fluoxetine or sertraline as first choices. This establishes an older local preference; it does not establish continuous application or unchanged wording in every later version. Search-index publication/crawl dates were not used as policy dates. Source: https://rightdecisions.scot.nhs.uk/media/1803/treatment-of-depression-pathway-final_.pdf 4. NICE CG113 recommendation 1.2.23: dated 2011, amended 2020. For generalised anxiety disorder, it already favoured considering sertraline first on cost-effectiveness grounds. This is diagnosis-specific NICE context, not proof of a new Scottish policy in 2024–2025. Cost-effectiveness includes benefits and costs and cannot be inferred from our mean ingredient cost per dispensing item. Source: https://www.nice.org.uk/guidance/cg113/chapter/Recommendations#drug-treatment 5. MHRA citalopram/escitalopram QT safety restrictions: article dated December 2011 (the GOV.UK page itself was published December 2014). These longstanding safety considerations predate both observed years. Their existence alone does not identify a cause of the annual changes. Source: https://www.gov.uk/drug-safety-update/citalopram-and-escitalopram-qt-interval-prolongation MONTHLY TIMING IN THE EXISTING DATA Each share uses the same month's national item total for all six SSRIs as its denominator. Sertraline: Jan 2024 48.826%; Jul 2024 49.669%; Dec 2024 49.925%; Dec 2025 51.116%. Citalopram: Jan 2024 23.607%; Jul 2024 23.178%; Dec 2024 22.862%; Dec 2025 21.808%. Before the guide, Jan–Jul 2024: sertraline +0.842 percentage points; citalopram -0.429 percentage points. These are endpoint descriptions, not fitted trends or causal estimates. Monthly variation remains visible in the chart. The two shares are not complements because four other SSRIs are included. FORMULATION CHECK USING EXISTING DETAILED FILES 2024 2025 Change Sertraline tablets 2,064,423 2,132,080 +67,657 Sertraline oral/liquid 2,013 3,894 +1,881 Total 2,066,436 2,135,974 +69,538 All 50mg/5ml suspension descriptions: 328 → 2,258 nationally; 111 → 589 in the three East Region dispensing areas combined. The extract does not carry a verified licensing-status field, so these are presentation matches, not counts proven to be the newly licensed product. All formulations reconcile to the original sertraline annual totals. No unclassified descriptions remain. GEOGRAPHIC ALIGNMENT 2024–2025 annual item changes in the East Region dispensing areas: NHS Borders: sertraline +4.91%; citalopram -4.70%. NHS Lothian: sertraline +9.34%; citalopram -6.81%. NHS Fife: sertraline +2.94%; citalopram -2.63%. The same opposing directions occur in all 14 territorial dispensing areas, including those outside this committee's three-board scope. Dispensing area is not prescribing-practice location, patient residence or a verified measure of exposure to a formulary decision. These are descriptive comparisons, not intervention and control groups. LIMITS AND USE IN THE COMPETITION This was a targeted review of official national guidance, dated local pathways and committee records, not a complete historical version audit of all 14 boards. Failure to find another change is not proof that none occurred. Current undated formulary wording was not treated as evidence of when a policy began. Draft consultations and post-2025 final guidance were not counted as implemented policies during this window. The monthly files measure paid dispensing items, not prescribing decisions, new starts, individual switches, diagnoses, adherence or outcomes. Payment timing, stock/pack choices, repeat intervals and calendar effects can affect counts. The figure uses monthly item shares to describe medicine mix, not a rate of mental illness or treatment success. With only 24 months, unknown implementation and no patient/prescriber linkage, we cannot attribute the pattern to the August guide, safety advice, costs or prescriber incentives. Recommended interpretation: “Sertraline gained and citalopram declined across dispensing areas. Guidance favouring sertraline existed before this period, and the monthly movement was already underway before the August 2024 national guide. The dispensing records do not tell us what caused the change.” Use guidance as explanatory context. Do not present a new 2025 policy, individual switching, better care or cost saving as the demonstrated finding. The current prototype treats guidance as context and does not attribute the observed changes to it. Nothing has been submitted; reader feedback and the user's explicit readiness instruction remain required before submission.